What is an associated conversion? Multi-channel sequences

” and how they can help qualitative analysis of your advertising campaigns. Today we will talk about what tools can be used for this, and how.

At the moment, the most popular web analytics systems in RuNet are Google Analytics and Yandex.Metrica (who doesn’t know yet, both of these tools are free). Both tools have support for multi-channel funnels, but they implement it differently. So let's look at what and how.

Let's start with Yandex.Metrica, because it implements this functionality quite simply. Moreover, for it to work, you also DO NOT need to update the YaMetrica html code or enable anything somewhere, everything is processed automatically by itself.

If you have goals set up, then in the “Sources - Summary”, “Sources - Search Engines” and “Sources - Advertising Systems” reports you should have a “Delayed Conversions” column:

That is, let’s calculate using the example of the screenshot above: there were a total of 54,630 visits from search engines, among which there were 203 targeted ones (that reached the goal). As a result, direct conversion (according to the latest source) was 0.37%, which is what we see in the report. But in addition to this, there were another 266 deferred conversions, whose visitors initially came to the site from a search, but then converted by visiting from another source. That is, we can say that the search influenced another 266 conversions, or sales, or registrations (depending on what your goals are set up for).

And the level of deferred conversion in Yandex Metrica is calculated as follows:

Data for deferred conversion is calculated only for the last two weeks
If the user has not been on the site for more than 30 days, then information about the original traffic source is deleted.
If the user has never been to the site and achieves the goal on the first visit, goal achievement is only counted in direct conversion. In deferred conversion, only goal achievements are taken into account when the user visits the site again.

As far as I know, this is all the current Yandex.Metrica functionality that concerns multi-channel sequences. With its help, you can evaluate what part of third-party conversions each source generates, and whether it’s worth disabling it if it doesn’t handle direct conversions well.

Now let's move on to Google Analytics, its support for multi-channel sequences is implemented much more widely, and there is even a separate category of reports dedicated to this, which is located in the “Conversions” section.

But before we talk about reporting, let's understand the mechanics of accounting for these “multi-channel” conversions in Google Analytics. Let me briefly remind you that the main concept here is that before making a conversion on your site, a visitor visits it several times from different sources (forming a chain of visits).

First interaction conversions(First Interaction Conversions) are those visits with conversions that originally came to the site from a given channel or source. That is, the current source was at the beginning of the chain of visits.
Associated conversions(Assisted Conversions) are those visits with conversions in which a given channel or source participated in the chain of visits. That is, it was either at the beginning of the chain or in the middle, but not at the end.
Conversions by last interaction(Last Interaction Conversions) are those visits with conversions for which this channel or source was at the end of the chain of visits. That is, it was the last one before the conversion on the site. The usual direct conversions are determined using the same scheme.

For clarity, I will show the process of accounting for conversions and determining their types using the following example:

Suppose a certain Vasily bought himself an apartment and now needs to equip it. He decided to start with the most important thing, namely to buy a refrigerator. And you are the owner of an online store of household appliances. But Vasily doesn’t know much about refrigerators, since he spent all his time earning an apartment. Therefore, before he comes to you, he needs to study what kind of refrigerators there are, and which one is best suited for him and his budget. Moreover, he must understand why buying online is better than offline. Only after this will he go looking for a suitable online store to purchase. By the way, if you really are the owner of an online store of household appliances, then check yourself, at which of these stages does the client first meet you? The sooner, the greater the chance of making a client yours.

At the stage of studying refrigerators, Vasily came across your store in a search and switched to it (moving from the search). After reading the site, he decided: “It’s a good choice, I’ll probably buy it here when I save up the money for a refrigerator.” After which he closed the site with a clear conscience and left. And over time, I completely forgot about the store, since there were so many other things to do. By the way, about 80% of all new visitors to most online stores do this. Because they know the main thing is that the product they need is on the Internet, and in which specific store it doesn’t matter, because then they can always find it through a search, if not this store, then another.
Owners of online stores, think about this: “Why should a visitor remember your website or store name if he can always find a similar seller in the search? How can you make sure that among all the options, he remembers and chooses you?”

After some time, Vasily decided on the model of the refrigerator, and since he had long forgotten about your store, he again went into the search to look for candidates. And lo and behold, he came across your ad from contextual advertising, and what’s more, it seemed attractive to him and he went to the site (transition from advertising).
After wandering around the site, he found the refrigerator he needed, but that’s not a problem, he doesn’t have the money yet. Therefore, he postponed the purchase until salary, and bookmarked the site so as not to forget.

Having received his salary, satisfied Vasily returned to your site through a bookmark (direct transition) and made the long-awaited purchase.

As a result, the complete chain of visits for this conversion (purchase of a refrigerator) will look like this: Transition from search -> Transition from context -> Direct transition. And in Google Analytics reports it will be shown like this:

At the same time, it is important to understand that the conversion to sale will be the same as the actual purchase. But for different sources it will be shown differently:

If we didn’t have this functionality, we would only see conversion from direct traffic. And so we know that in fact, those who initially came to the site from other sources convert through direct traffic.

And on the other hand, if we see that paid advertising does not pay off and provides few direct conversions, then we need to check whether it delivers conversions for other channels?

Multi-Channel Funnel Reporting in Google Analytics

We've sorted this out, now let's see what reports Google Analytics provides us with this functionality. Let me remind you that they are located in the main menu on the left, in the “Conversions” category, subsection “Multi-channel sequences”.

Before examining these reports, pay attention to the filter that is located above each of them:

Here you can select which conversions you want to consider in the report. You can select conversions for all goals, or just one. You can also choose the type of traffic you will consider: traffic from all sources, or only from AdWords. For some reports, you can additionally select the length of the sequences for which data will be shown.

As for the reports themselves, the first of them is Overview report:

The chart shows how current conversions (sales, registrations, downloads, and other goals you set up) are distributed across the main channel groups. The report is more suitable for quick, superficial analysis.

You can select any 4 channel groups to display on the diagram. If you do not hover your mouse over the chart itself or its zones, it will be shown what proportion of conversions fall into the specified channel or specific intersection.

For example, in this screenshot we see that during the specified period, 27 of all conversions (3.66%) were made by visitors whose chain included all 3 selected channels.

Assisted conversions report

This report shows detailed statistics on conversion types for individual channels and groups of channels (1). For each of them, the number of direct and associated conversions is displayed, and, accordingly, their value (amount of orders or goal value). Also, for a more detailed study, in this table, instead of channel groups, you can display specific traffic sources or keywords; for this, use the menu above the table (4).

The coefficient in the last column (2) shows the source’s “propensity” for a certain type of conversion. That is, it shows the nature of the channel’s operation.

Namely, if this indicator is less than 1, then this channel or source is more often found at the end of the chain of visits. That is, it often works as a closing source and gives more direct conversions than associated ones.
And if this indicator is greater than 1, then this channel or source is more often found at the beginning or middle of the chain of visits. That is, it works more as an initiator or support of conversions.

At the top, this report can be switched to another view (3), which will show the exact number of conversions for the first and last interaction.

Top Conversion Funnels report

But this is perhaps the most curious and interesting report, because in it you can see the specific sequences of traffic sources from which visitors come before making a conversion.
The principle of generating a report is the same as the previous one. You can choose to display groups of channels as well as specific source sites or keywords. The resulting table can be filtered, which is very convenient for analyzing specific sources.

For example, you can select all occurrences of a particular site or keyword and understand at what point the visitor accessed it. You can see how the visitor specified search queries in the search engine when arriving at your site.

Please note an important point: when you first launch these reports, Google Analytics generates the basic groups of channels (which are shown in the screenshot) at its discretion, based on current sources. Therefore, the composition of these groups may differ from what you actually have, which will lead to incorrect display of data. For example, in the “Social Networks” group he writes about 150 of the most famous and popular social networks known to Google (VKontakte and Odnoklassniki are there). But it is quite possible that among them there will not be those that your audience uses. The same applies to the “Paid advertising” group; only clicks on links with certain tags fall into it. If your tags are different, then the transitions will not fall into this group. Here's a list of what Google analytics lists in basic groups.

Therefore, it is important to check the groups for compliance with your realities and make changes; this can be done in the menu above the table. In the Channel Groups section, select Copy Basic Channel Group Template. A list of current channel groups will open in front of you, in which you can either make edits to an existing group or create a new group based on the old one. The principle of operation is the same as for user segments.

After editing, these groups will appear in all reports in the “Multi-channel funnels” section, including the first one with a convenient diagram.

Time to Conversion report

This report shows how many days pass from the first visit to the site until the conversion to the specified goal. That is, you can find out how much time a visitor spends on average making a purchase decision and going through the entire chain of visits. For example, the third row of the table in the screenshot tells us that for 21 conversions, the time it took to complete the chain of visits was 2 days.

Conversion path length report

This report shows the number of conversions by clickstream length. For example, the third row of the table in the screenshot tells us that 73 conversions were completed after the visitor went through a chain of 3 elements.

These are the rich reports Google Analytics provides for multi-channel funnels. So use it and study the behavior of your audience, I’m sure you’ll learn a lot of new things.

Well, in the next article in this series, I will talk about specific techniques for analyzing advertising campaigns using these reports.

If you've read this far, I want to thank you for your patience and praise you for your perseverance. I am sure it was not in vain, and you gained useful knowledge. The article really turned out to be long and it took a lot of time to write.

“Commander, fill me up with a full tank, and I’ll go!”

This is how they buy gasoline. In a similar way - bread, chewing gum, milk. The need arose - I went and bought it. These are simple and cheap goods (even gasoline; sorry). But that’s not how they buy rocket fuel or burglar-resistant doors. A potential buyer takes a long time to choose them and compares competitors’ offers. This is the specificity of complex and expensive products. And this is our area of interest. Completo marketers work with B2B and businesses in complex markets.

To sell something complex, a company takes the potential buyer through the customer journey. To do this, it uses marketing channels. How to determine the most effective of them is discussed in the article.

There are many canals, like in Venice

How to analyze multi-channel funnels. Analytics reports

Path to reports: conversions → multi-channel

Report 1: Sequence Length
Shows how many times a user visits a site before taking a target action.

Set parameters:

Select a conversion.
Type is everything.
The type of interaction is everything.
Days before conversion - maximum 90.

In the example, we selected only one type of conversion - transactions.

Here, 42% of purchases occur on the first visit to the site. The remaining 58% - after several visits:

For most businesses and for purchase conversions, these are typical numbers. For free actions the statistics are different. Of those who sign up for email newsletters or download free content, 90% do so the first time they visit a site.

If more than 90% of your conversions happen on the first visit, you don't need multi-channel funnels.

Report 2. Main conversion paths

The report shows which sequence of advertising channels brings the most profit:

Use the best connections.

Report 3. Time to conversion

The report shows how many days pass from the first interaction with the site to the conversion:

Here, 62% of conversions happen on the first day.

Use this data to determine an effective period for retargeting. For example, if there are few or no conversions within 15 days of launching a campaign, you should remove users from your list.

Or we see a surge in activity occur on the seventh day after the first visit. Then we set up retargeting specifically for these users.

Report 4. Assisted conversions

Here are statistics on assisted interactions, first interactions and conversions. We are interested in the first two categories. From them we will find out which channels are the main ones and which are auxiliary.

- Analysis of assisted conversions

For each channel the following are displayed here:

Number of associated conversions. These are those in which the channel played a supporting role.
Number of conversions based on last click or direct interaction.
The value of both types of conversions, broken down by channel, if the selected conversion value is configured to be sent to Google Analytics.
The ratio of assisted conversions to last click or direct interaction conversions.

The higher the number in the last column, the less role the channel plays. A value greater than 1 is typical for unimportant channels. They are present in the customer journey, but conversions occur through referrals from other channels. For main channels this ratio is less than 1.

In the screenshot we see that “(direct) / (none)” is the closing channel. Direct and AdWords are auxiliary; They are less likely to bring conversions, but they participate in the entire chain of interactions.

IMPORTANT! When completed conversions are reported to Google Analytics using the Measurement Protocol, the last source or channel in the report will always be “direct / none”. This is a feature of how Google Analytics works with the Measurement Protocol.

The second tab shows the number of conversions through channels after the first interaction with them:

The report is useful for analyzing branding campaigns or campaigns that promote a new product. Use it to find out which channels and advertising campaigns are connecting with users.

Direct is in the lead in the screenshot. Without this channel there would be no conversions. We know that this channel is auxiliary, not closing. If we assessed the effectiveness of Direct only by last click conversions, we would make the wrong decision - cut the budget or abandon the channel.

If you analyze a report where channels are grouped by default, you may draw wrong conclusions. For example, the value “yandex-direct / cpc” of the “Source or channel” parameter contains data on advertising in search, YAN and retargeting. These are different sources, so the report for this group of channels is the average temperature for the hospital.

For correct analysis, create groups of channels.

Example
We know that the “yandex-direct” source includes campaigns for thematic keywords, brand name and YAN campaigns. To evaluate the effectiveness of sources, we break down “yandex-direct” into:

For this:

1. Go to the admin panel

2. Select: Channel settings → Channel group

3. Click on: New channel group

4. Enter the name of the channel group and set the conditions for YAN, search and brand. To search - for example, like this:

5. When analyzing associated conversions, select the channel group that you just created in the report:

Now the influence of each source is clearly visible:

Divide the channels - “Brand”, “Search” and “YAN” - into groups, and groups into campaigns. This will further increase your accuracy.

Would you like to receive an offer from us?

Start cooperation

Attribution models and budget reallocation

The attribution model in Google Analytics is the principle by which value is allocated among channels in the conversion path. A good attribution model shows how much revenue each channel generated. Next, determine ROMI - and you will see the value of each channel. Now you know which channels to allocate your budget to.

Attribution models

By default, Google Analytics uses the Last Indirect Engagement model. Read descriptions of all standard models.

You can create your own based on standard attribution models.

Tab “Administrator” → Attribution models:

But you can also use one of the standard models.

For our example, the “Position-linked attribution” model is suitable for us, so we will not create a new one.

IMPORTANT! You cannot compare different channels with each other. The user who came to the site based on a branded request and the one who clicked on the banner in the Display Display are different people at different stages of decision-making. Therefore, it is incorrect to conduct analysis using the “Last interaction” attribution model.

Determining the value of the channel and redistributing the budget

How do you determine the value of each channel in terms of ROI and marketing budget allocation?

Suppose the marketing budget is 1,000,000 rubles per month, and is distributed as follows:

As we said above, in the example under consideration, the “Position Based” attribution model is completely suitable for us. But nothing stops you from choosing yours.
We are interested in the “Conversion Value” column:

From this column we see that the income from each channel can be distributed like this:

Yandex.Direct: 700,000 / 300,000 * 100% = 233%
Google AdWords: 300,000 / 200,000 * 100% = 150%
VKontakte target: 270,000 / 170,000 * 100% = 159%
Facebook Target: 250,000 / 150,000 * 100% = 167%
MyTarget: 190,000 / 180,000 * 100% = 106%

Now we determine new budgets for each channel. To do this, we calculate the channels’ shares in profit (% of total profit). Adding up all the income, we get: 1,710,000 rubles. Of them:

Yandex.Direct - 41%
Google AdWords - 18%
Target VKontakte - 16%
Facebook target - 14%
MyTarget - 11%

Distribute the budget in the same shares:

Channel	Old budget rubles	New budget rubles
Yandex.Direct	300 000	410 000
Google AdWords	200 000	180 000
Target VKontakte	170 000	160 000
Facebook Target	150 000	140 000
MyTarget	180 000	110 000

Uploading data using the API

You can automate this process. To do this, set up uploading data from the Google Analytics API to Google Spreadsheets and write down the formulas.

Create an empty table and install the Google Analytics extension.

Create an empty report:

A template will appear:

Line 4 contains the view ID. Data will be downloaded from it. Other parameters are described in the instructions for developers.

Fill in the template with parameters. Check the syntax using Query Explorer. In the menu, click on Run reports. You will receive the download on a separate sheet.

After setting up the formulas, a plan for distributing the budget across channels for the coming period will appear.

Success! Share with friends and colleagues

Happened? We hope so. Share your opinion about this tool and our instructions in the comments. Share the link to the article with your friends and colleagues. But: under no circumstances with competitors.

(Visited 2,521 times, 1 visits today)

Sometimes a lot of time passes from the client’s first acquaintance with you to the purchase. To reduce it, you use advertising. An attribution model and multi-channel funnels help you monitor the effectiveness of your advertising channels.

In this article you will learn how to determine the effectiveness of advertising in Google Analytics and audit conversion channels.

Basic attribution models

They accurately determine the source of referrals to the site. The report in Google Analytics is easy to customize to suit your needs.

There are five basic models:

1) Attribution based on the first interaction (click). The value is assigned to the source that brought the visitor to the site for the first time. The model is useful if the purpose of advertising is to arouse interest in a product or site.

2) Attribution based on the last interaction (click). The value is assigned to the last channel (their total number can be any) before the conversion. The model is focused on direct transactions and is therefore considered basic.

3) Attribution of the linear model. Each channel is equally valuable throughout the chain. For example, if a customer went through four channels before converting, everyone gets 25%. The model is suitable for assessing ongoing interactions with customers.

4) Positional attribution model. A combination of the first two models. The main values are divided between the first and last channels. For example, the first is 40%, the second is 20%. The latter also receives 40%. What is important here is not only familiarity with the product/site, but also the conversion channel.

5) Attribution of the recency of the interaction. A simple (and therefore popular) algorithm, used mainly for short-term or one-time promotions. The main value is assigned to the channels that “shot” during the promotion.

The path to conversion is the sequence of steps a visitor takes before conversion. Reports in the “Conversion” section, “Multi-channel sequences” subsection will show how much time it takes to complete a target action (call, request, order).

Assisted conversions report

It shows where visitors are coming from.

Pay attention to the last point. If the number is less than 1, this channel is usually the last one. More than 1 - it is more often found at the beginning or middle of the chain of visits, initiates or supports the target action.

The report also shows statistics by conversion type for individual channels and groups of channels. Only 2% of visitors are ready to buy on their first visit. Most people leave the site after 15 seconds: they read reviews, study the seller’s accounts on social networks - this brings them closer to conversion.

Top Conversion Paths report

The report will show what visitors do along the entire journey. You see the following sequence of traffic sources:

Also, these are the conversion paths for a group of channels:

If referrals appear in groups, study the traffic from referral sites and find the links that brought traffic.

Customers find you through search, but advertising convinces them to buy. If the “conversation” starts with PPC (Yandex.Direct & Google AdWords context, targeting in social networks), and organic search leads to conversion, think about whether you are spending money on that.

Time to Conversion report

The report shows how many days pass from the first visit to the conversion - the time until the user considers the purchase decision.

If 50% convert in 12 days or more, spend time nurturing customers. Improve your content or create it for different devices to keep visitors coming back. Start a newsletter to remind yourself. Work until the number of days in the report is reduced.

Sequence Length report

Shows the number of conversions along the chain of visits:

For example, from the second line you learn that 37,517 target actions occurred after passing a chain of two elements.

Study conversion paths to break down incoming traffic. Compare criteria: mobile vs desktop, new visitors vs returning clients.

1) Customize models to suit your goals, specific platform and audience.

2) Use auto-tagging in Adwords and other tools and UTM tags for social media campaigns.

3) Consider customer life cycle (LTV). A good customer comes back again, which is why he is 18 times more expensive than the average customer. “Old people” are more likely to return through direct traffic (from bookmarks), social networks (learn about sales), and email (learn about promotions and discounts).

New customers usually come through paid advertising, organic search, referrals and social media.

It is twice as expensive to retain a customer than to acquire a new one.

High conversions to you!

February 14, 2018

Configured goals/conversions (“Achieved goals” inGoogle Analytics) – these are the main indicators of the site’s effectiveness. Thanks to them, you can find out how the actions of visitors affect the sales of your products and services.

Using goals and (for online stores), you can determine income, the number of transactions, the average bill, make a list of best-selling products, and track the number of received applications and requests. And look at all this in the context of each attraction channel.

Staging is one of the priority tasks for any business. This is why it is so important to choose the right metrics to track before promoting a website (money costs).

Category reports "Conversions" store comprehensive data about the user’s actions and consist of several blocks : “Goals”, “E-commerce”, “Multi-channel funnels” And "Attribution".

Goals

shows a summary of the achievement of goals with the main indicators inherent in this section:

Achieved goals– total number of conversions;
Goal value– the total value of all conversions in monetary terms for the selected period. If it is not specified in any of the goals, then the total value will be equal to 0;
Goal Conversion Rate— the ratio of the number of executions of target actions to the number of sessions;
Overall interruption ratio– number of interrupted targets. Defined as the total number of aborted sequences divided by the total number of goal starts;
Statistics for each goal separately.

In the dropdown list "Target" You can choose any of them to analyze the results:

In the report « URLgoals" we can see the list of pages on which this conversion took place:

The example above shows pages with a dynamic ID parameter, each of which corresponds to a separate order in the online store. The goal value was not specified during setup.

contains information about the pages preceding the final goal.

Thanks to it, we can see the user’s path through the last 3 steps before conversion: what pages he visited, what he viewed, etc. As a rule, the steps preceding ordering in an online store are always the same: Cart -> Checkout -> Delivery and payment -> Order.

Based on this presentation of data, it is easy to find pages where users “fall off”. And what was the proportion of their total number.

It may be that you have a large number of fields to fill out on your website and this causes inconvenience to potential clients. Or there is no way to remove unnecessary goods from the cart, so the user leaves it without ordering anything.

By identifying bottlenecks on your site, you can quickly make changes to increase conversion at each step of the funnel.

"Goal Map" very similar in functionality and presentation to the event map and the classroom.

It allows you to answer a number of questions related to user behavior and their interaction with the site:

Do they go through the entire conversion path or just part of it?
On which pages does the sequence most often break and why?
Why do users go back a step and what has become unclear to them?

In the example above, some users move from one step "Order placed" one back. Most likely, they wanted to change something in the order information and decided to come back to correct the delivery time or correct the destination address. Or from the form "Place an order", when some clients returned to the stage "Item in cart" to add/remove items or enter a discount coupon.

Below the path map, a table is displayed with additional information on the selected funnel indicator. Sources can be compared by any other indicator. For example, by selecting the indicator "Campaign", you can find out which advertising campaigns were most effective and which should be optimized or disabled altogether.

You can also apply and analyze statistics on non-standard traffic types to the report.

Google Analytics has several conversion segments available by default:

Visitors who didn't convert– all sessions of users who did not make a single conversion during the reporting period;
Visitors who converted– all user sessions that completed at least one conversion in any of their sessions during the reporting period.
Sessions with conversions— segment for sessions during which at least one goal achievement was recorded.

To evaluate the performance of smart goals, Analytics provides a corresponding report. We discussed in the article what information it contains and how to use it in practice. "Smart Goals".

You can view the main data for each available conversion (including all goals at once) in any standard report, for example, in "Traffic Source - Source/Channel".

"Traffic Source - Source/Channel" report

Electronic commerce

This group of reports is designed to analyze customer behavior. It contains information about products and transactions, revenue for each of them, average order value, transaction rate, time to purchase and other indicators.

To make e-commerce data available in Google Analytics, you need to enable it in the view settings, and also add data collection code to the site. Read more about this in this article.

Important: In reports, income will be displayed in the currency whose icon is indicated in your view. Therefore, if you sell goods in rubles or hryvnias, and the report contains the icon "dollar", you need to change the currency settings in this view.

The global currency (the one that is the default on your site) is used for all transactions and products. If your site allows you to pay in multiple currencies, then using an ecommerce plugin you can specify the currency to use for the transaction.

The local currency must be specified according to the ISO 4217 standard. For a complete list of conversion currencies available in Google Analytics, see Google's Currency Codes Reference Guide. The procedure is changing the tracking code on the site using the counter property currencyCode.

allows you to get acquainted with the main indicators of the online store and general information on products, including the best-selling products.

The general purchasing process that every user of our site goes through is presented in.

It consists of 5 stages:

All sessions
Product viewing sessions
Sessions with adding items to cart
Checkout sessions
Sessions with transactions

By default, the report is built in the context of sessions and user types - new and returning. Thus, by selecting the required parameter (source or channel, device type, etc.), we can, in addition to setting up regular goals, return dropped visitors at any step of the sequence.

If, when analyzing customer behavior, it turns out that they leave the product card page without adding it to the cart, then most likely:

visitors were looking for another product;
they simultaneously compared several sites and found the best offer from competitors;
there was a glitch on the page and the person simply couldn’t do it.

If visitors leave the site after adding to cart, this may mean that:

they found a better offer;
they were unable to complete the order for some reason not related to us (the Internet was turned off, someone was distracted, they just got too hungry);
There was a glitch on this page and the person was unable to complete the purchase.

If customers leave the site already at the payment stage, then in 90% of cases this is due to:

with an overly complex procedure;
with the fact that there is no suitable method (for example, only online payment or only cash to the courier);
high delivery costs.

Depending on the tasks, we can choose any of the available options.

By clicking on one of the blocks, we will be asked to create an extended segment, which can later be applied to reports or created remarketing lists based on it.

To do this, simply import the segment from property-level audiences into your AdWords account.

Switching the table to "Interrupts", we will have access to information on unpaid carts, unfinished purchases and sessions in which no items were added to the cart.

Using the report, you can determine at what stage visitors most often leave the site, how much you could potentially earn from them, and how new and returning users behave. It also lets you know which pages are causing the most churn, what devices this is happening from, and what next steps you need to take to correct the situation.

Remember in behavior reports we talked about quality characteristics and endlessly improving your site through page testing, design changes, usability, etc.? In e-commerce reports, this is especially important because you see the “weaknesses” not only in the form of beautiful graphs, but also in monetary terms. Don’t forget to work through your audience, creating remarketing lists for users with a high abandonment rate and those who have already taken one step forward and added a product to their cart.

In the report "Analysis of behavior when making a purchase" You can track user behavior throughout the entire purchase process and identify problem areas on the site. In some ways it is similar to the previous one, but there are still differences. Here we create the sequence of steps ourselves.

Google Analytics determines funnel steps based on the tags you set up when you initially set up ecommerce (Checkout Step Tagging).

The first three stages ( Item in cart, Delivery, Payment) are those that were created manually in the e-commerce settings. Last ( Sessions with transactions) is the number of sessions during which the transaction was completed.

If you have a suspiciously large difference between the stages, then most likely these steps are incomprehensible and intimidating for users. For example, you should not offer visitors only pay for goods online, without the right to choose other types. This will completely kill your business and you will never build a customer base or capture the market. Only very large companies that have hundreds and thousands of loyal customers, and their brand is constantly heard and trusted, can afford to experiment like this.

Difference between steps "Item in cart" And "Place an order" may also indicate that users do not like the large number of fields to be filled in (name, phone, email, address, your favorite musician, pet name, etc.) and they leave this page due to so much meaningless information. Try to make order forms very simple and understandable. A name and phone number are more than enough at the first stage of contact. All other information can be found out during one-on-one communication with the client.

Product effectiveness

Product Performance Report

In this report, you can see a general summary of the number of sales of a particular product, income from the product, number of transactions, average check and return amount.

Additional indicators of buyer behavior:

Selected Products Ratio— the number of product additions to the cart divided by the number of product information views.
Completed Purchase Rate— the number of unique purchases divided by the number of views of pages with information about the product.

Main parameter: Product, Product ID, Product Category (Advanced Ecommerce) And "Product brand".

In addition, in the report on the graph you can compare two necessary indicators from the proposed ones, and by switching to the tab, evaluate the following metrics for each product:

number of views of products from the list (how many times the product appeared in the list of products);
number of product information views (how many times users viewed the product information page);
number of product removals from the cart;
number of completed product purchases (how many times this product was included in the checkout process);
number of unique purchases (total number of transactions involving the specified product).

Thanks to the report "Product Efficiency" Can:

find products with a high number of purchases and additions to cart. They can be placed in more effective positions in the online store (on the main page) or you can create a separate banner for them so that potential customers immediately see such products and they catch their eye;
Identify products that are often viewed but rarely purchased. Apparently, the user found a similar product in another store at a more attractive price. You can make a discount on such goods or change the price, making it competitive;
identify products that are secondary(which add to the main ones) and add them to the recommended block under the card main product or on the ordering page;
analyze the popularity of certain categories on your website, which will allow you to redistribute budgets when purchasing goods in the future;
by adding data on the marginality of each product or product category, determine the price that can be assigned to the product to obtain greater profits. Perhaps you priced the product too low and it could have been purchased at a higher price.

Report on "Sales Efficiency" displays each transaction by its ID. Having configured the e-commerce module for an online store, you can see all order IDs not only in the site admin panel, but also in this report.

When you click on a transaction, you will see data on this order: the name of the product, the number of goods purchased and the income from them.

You can also transfer shipping and tax amounts and, in case of a return/cancellation of a purchase, import data on these transactions. An example of a return is discussed in detail in Chapter 5 “Resource”.

All pages and blocks on which you can view products are called product lists. They are generated by Google Analytics based on plugin tags ec.js.

You can analyze your own product listings: which of them is the most viewed, which of them has the most products, and which items in this list are the most popular.

In the screenshot above, the most viewed item was the product list. "Catalog", which had a click-through rate of 4.38%. It is also the most profitable for us compared to other lists. However, we see that the list of products "New Year" has the highest CTR (8.41%). This is explained by the time of year (the material was written in December 2017) and the interest of buyers in a specific product.

By going inside the list, we have access to information on each product:

how many product views there were;
how many times you clicked on products from the list;
Product list CTR;
number of items added to cart;
how many unique purchases there were;

Main parameter: “Product list title”, “Product list position”, “Product” And "Product ID".

E-commerce metrics are also presented in the tab "Statistics" In chapter "Electronic commerce" in many standard Google Analytics reports.

When you configure the Enhanced Ecommerce module, a tab becomes available in ecommerce reports "Marketing", which includes the following reports: “Internal Campaign”, “Order Coupon”, “Product Coupon” And "Partner Coupon".

The first report provides us with data on the number of views of advertising banners on our website, clicks and click-through rate (CTR).

Internal ad impressions are recorded on page load and transmitted with the initial screen view using the command ec:addPromo. In order for the data to be displayed in this report, you need to change the tracking code by adding additional variable values: promotion ID, title, material data and its position on the site ( id, name, creative, position):

Main parameter: "Internal campaign name"(this name variable in the tracking code is a string).

Using this report, you can determine which banner brought you the most income, which banner with which call to action has the best CTR, which of them arouse interest among visitors, but do not make purchases, and when you need to change the banner on the site due to a sharp decline demand.

The report on advertising campaigns on the site can be effectively analyzed with the report. It represents the queries that site visitors enter when searching for a particular product on your site. Thanks to reports And "Search queries"(on the website) you can identify the most popular products both in terms of internal search and clickability on them, and then place them in recommended blocks or favorites, in-demand products, etc.

Order Coupon, Product Coupon, Partnership Code

Analytics allows you to track product and order coupon conversions, as well as affiliate marketing transactions. This information is presented in three reports:

Order coupon
Product coupon

If you have the opportunity to provide a coupon to site visitors for some kind of discount, you can further analyze the effectiveness of each of them.

For example:

New Year's sale (-30% on all products);
Cyber Monday (-50% on category “A” products);
Tatiana's Day (-80% on all products for customers named Tatiana);
savings system for regular customers (-3%, -5%, -7% after each order);
etc.

The coupon can be applied to a specific product or to the entire order. In the first option, statistics will be available in the report "Product coupon", in the second, the data will be included in the report after 24 hours "Order Coupon".

The last report is designed to collect data on partners and affiliates. Statistics will only be collected if key values ( affiliation default)

It is rarely used, but it can be useful for those who have given their goods for sale to several online stores (or created several of their own on different domains), and record all data for all partners under one account.

Multi-channel sequences

If you are promoting a website to one of your customers and observe a situation in which the cost of attracting a client turns out to be too expensive or there are no requests from the source, do not get upset and do not make hasty conclusions. Now is the time to take a look at your cross-channel funnels and associated conversions reports.

Throughout this book, I gave several examples of how clients decide before purchasing how many sources they can go through before you see a completed application/order in your admin panel or in the mail.

For example, a person was looking for something on the Internet and the first interaction with your site was through Google AdWords contextual advertising (position 1). He then left the site and began comparing similar offers with others. After some time, he enters a query in the Google search engine and sees your site again, only now he goes to it through organic search (position 2).

- “Oh, a familiar site! I've been here before and it seems pretty good. I'm happy with everything: the price, delivery conditions, and it looks pretty nice. We need to write down the name so we can quickly place an order tomorrow.”

The next day, this person visits the site (position 3) and makes a purchase (position 4). The last interaction is a direct entry (direct / none).

What did we get?

Three interactions: the first is Google AdWords (cpc), the second is Google (organic), the third and last is direct entry (direct);
Conversion sequence length: 3;
Time to conversion (in days): 2 days.

In standard Google Analytics reports, all conversions are attributed to the last source, with the exception of direct traffic (direct / none). Revisiting Attribution Models and Chapter 4 "Performance". Thus, in our example, in the reports we will see the conversion of the second interaction (organic).

Interactions 1 and 2 are auxiliary (associated). But it would be wrong not to take them into account. After all, the first touch occurred through paid advertising, for which we paid money. And it was thanks to her that the entire chain of interactions between a potential client and the site, who later became our customer, began.

There are situations when users click on a campaign "Competitors" to the site, get acquainted with it, and then look for reviews on the forums. After reading, they move from the forum to the website and make a purchase. Referral traffic was the last in the chain of interactions, but if there had been no paid advertising for competitors, then there would have been no order in principle, since the client would have no way of knowing about our existence.

Multi-channel funnel reports help you evaluate the supporting contributions of different sources and make better decisions about their effectiveness. They are created based on conversion paths, which is the history of interactions (clicks or conversions) that led to a transaction on your site.

Main features of reports:

To analyze the received data, it is necessary to set up goals or e-commerce;
By default, only data from the last 30 days before the conversion is taken into account, but this period can be changed from 1 to 90 days using the switch "Lookback Window";

data in the report arrives with a delay of 24-48 hours;

It displays an overview of the number of conversions and associated conversions.

At the top of the graph, we can select specific goals for which we want to analyze the data.

It is also possible to select the type of sources (All or AdWords) and change the lookback window.

Lookback Window (Attribution Window)- this is the period of time that can pass between the source and the conversion, after which the purchase will no longer relate to this source.

If we know that the user takes a long time to select a product and make a purchasing decision, then the lookback window when analyzing multi-channel funnels can be set to 90 days. If, on the contrary, the sales cycle is very fast, the attribution window value is set to minimum or left at the default of 30 days.

Total number of conversions is the sum of all goals and transactions for the selected period of time.

– these are conversions for which this traffic source was present in the chain of transitions, but was not the last one in it.

In the example above, users made 711 purchases, of which 414 were purchases where the customer visited our site at least twice in the last 30 days before the transaction before making the purchase.

Multi-Channel Conversion Visualizer ( Euler-Venn diagram) makes it possible to evaluate the role of each source on the path to the transaction. For example, from the screenshot above 56,42% conversions are made using a direct source, 34,69% - using free search, and 27,08% - thanks to search advertising. As part of our transitions 2,7% transactions are made as a result of user interaction through three sources: direct, free search and search advertising.

Euler-Venn diagram

Adding up all the sources, we see that the percentage of the total number of conversions is more than 100%. This is due to the fact that one channel could be in the chain either at the beginning, in the middle or at the end. That is, there are intersections of several channels.

In addition, Analytics provides the ability to select several conversion segments for simultaneous comparison (no more than 4):

For example, you can create a conversion segment that only includes funnels where the first interaction was a conversion of a specific value. Once applied, the segment remains active as you navigate to other multi-channel funnel reports. To return to viewing all conversion funnels, select a segment "All conversions".

We may remove or create our own conversion segments. To do this, when selecting conversion segments, in the upper right corner next to custom segments, click “Create a conversion segment”.

contains information about conversions where more than 1 source was recorded before the transaction was completed. That is, the user interacted with your site at least twice.

In 58% of transactions on our site, transactions were made from two or more sources. A very important indicator is , which in our example is 0.58 (58%).

If the number is greater than 1 (100%), this means that users take a long time and carefully to make a decision before making a transaction, and for our site the role of multi-channel sequences is very important. If this indicator is 1, then the number of regular and associated conversions is equal. If the indicator is less than 1 or tends to 0, then our business is characterized by purchases "Here and now", as users make decisions quickly.

1 < — more often acts as an auxiliary (associated) channel;
=1 - both;
1 > — often acts as the last (closing) channel.

In the report table, opposite each channel group, this number is presented in the last column. For example, search advertising has a value of 0.68, the smallest among all others. This tells us that contextual advertising users tend to buy immediately, without additional interactions.

In one of its webinars several years ago (more precisely in 2014), Google provided statistics on where an advertising network channel (display advertising, teaser networks, CPA networks, etc.) had value “Assisted/Last Click/Direct Conversions” = 17.

This research shows that ad networks have a huge influence on sales through other channels, but they themselves make virtually no sales. Recall the situation with the Display Network (Display Network) in AdWords: in all standard Analytics reports, these advertising campaigns have very expensive leads or no leads at all. And this is precisely the point - the source is auxiliary and it is not recommended to exclude it when promoting a site, since it affects other channels. The same channel is referral transition from other sites.

The column displays the sum of all chains of interactions in which this channel participated. And if we sum up all the associated conversions across channel groups, the result will be more than our 414. This is due to the fact that within the same chain we could have both a direct transition and, for example, a free search. The channels present in assisted conversions are not mutually exclusive—if two channels played a supporting role in the same path, the assisted conversion is counted for each of them.

The total number of direct conversions and transactions in the Multi-Channel Funnels report and other Google Analytics reports should be the same.

The value of assisted conversions is the income from all conversions made using this channel.

Value of conversions based on last click or direct interaction– this is the income from all conversions on the way to which the channel was the last interaction.

Main parameter : "Channel Group for Multi-Channel Sequences", "Default Channel Group", "Source or Channel", "Source", "Channel" or "Channel groups", which we can create manually.

You can also select an additional parameter from categories "Traffic Sources", "Users" And « AdWords» .

On a line graph we can display the values of the selected indicator for three variables:

by date of conversion;
by the number of days before conversion;
by position in the journey (by the number of interactions before conversion).

The report also includes tabs for assisted interactions, first interactions, and organic conversions.

contains a set of chains (conversion paths) for multi-channel sequences and information about which sources most often the user visited our website before making a transaction.

For example, for 31 transactions, users followed the path Free search -> Direct entry within 30 days before the transaction. Or, for example, 18 conversions were made through two visits, in which both the first and second visits to the site were Direct approach.

Notes x2, x52, x3, x69 in the screenshot means repeating the path in a row.

At the top of the graphs, we can set the funnel length, which shows how many days and interactions it took for a visitor to convert.

By adding an additional parameter "Campaign Path (or Source/Channel)" and by applying an advanced filter to the conversion paths of a group of channels, we can see by utm tags and ad names what role our campaigns played in the sequence.

3 transactions were completed within the path Search advertising -> Direct entry, and the campaign was branded. Multiple paths contain a chain Search advertising -> Free search and vice versa, Free search –> Search advertising.

In the same study, Google employees presented another slide in which it was stated that chains with a number of steps of 2 or more provide 1/3 of all revenues. There were 9 of them and they look like this:

Free search -> Free search
E-mail newsletter -> E-mail newsletter
Contextual advertising -> Contextual advertising
E-mail newsletter -> E-mail newsletter -> E-mail newsletter
Free search -> contextual advertising
Contextual advertising -> Free search
Free search -> E-mail newsletter
Contextual advertising -> E-mail newsletter
Advertising Network -> Free Search

The chain works best Free search -> Free search And E-mail newsletter -> E-mail newsletter, since in the first case the user independently finds information in the search engine without any intrusiveness on our part and decides for himself whether he should get acquainted with our offer or not.

The mailing chain works well because the user is loyal to us, he subscribed to it voluntarily, thereby agreeing to send useful materials and unique offers with a certain frequency. And when he receives such a letter with a promotion or discount, he immediately places an order with us without hesitation.

The interesting thing about the histogram is that the chains Free search -> Contextual advertising And Contextual advertising -> Free search work equally well. Therefore, to answer the question: what is better to invest in – SEO or context? We respond to the customer based on this data.

Thus, it is necessary to use these traffic acquisition channels and combine them with each other.

Note: research data on slides for 2014, the situation may have changed somewhat in the direction of social networks and instant messengers, but only slightly.

Main parameter: Channel Group Conversion Path for Multi-Channel Funnels, Default Channel Group Conversion Sequence, Source/Channel Path, Source Sequence, Channel Sequence And "Channel groups"

Important! Don't add up regular and associated conversions, as you risk counting the same order twice. It will seem to you that this source works well, and redistribute the advertising budget in its direction. But due to "twin" data, you simply incorrectly identified a profitable channel and lost money.

If you have no idea how long it takes a user to make a purchasing decision from the first interaction with the site to the final transaction, you can use the report.

As an example, let’s look at the statistics of an online store selling flowers. The table shows that 56.4% of all conversions occur on the first day of entry. The remaining 43.6% thought for some time. 18 conversions (2.53%) occurred the next day, 6 transactions (0.84%) two days later, 16 purchases (2.25%) three days later, etc.

It would seem that we should receive the main income during the first few days and the role of multi-channel for a flower shop is not so important, since this topic falls under the concept "Here and now". However, more than 30.1% purchase within 12-30 days of first interacting with the site. In this case, it is necessary to use additional forces in the form of remarketing campaigns and warm up users along the entire funnel path.

This may also include repeat purchases. Even if they were performed by the same user, they will still belong to different paths - the system does not count unique users, but the sessions in which they achieved conversion.

There are often cases when, within one day (time before conversion “0”), a user visits the site several times and places an order after some time. For example, on the 3rd visit. Analytics assigns all these sessions to such a visitor.

In the report on "Sequence length" indicates how many conversions occurred in conversion paths containing from 1 to 12 or more interactions.

You can use this report to analyze your audience and use it to determine at what length of the funnel they decide to convert. In other words, how many sessions does it need before executing a transaction? More than 40% (41.77%) of online flower store users make a decision in the first session. There is also a part of the audience (more than 25%) who are picky about purchasing a particular bouquet and need more than 12 interactions to complete a transaction.

Attribution - Comparison Tool

This tool allows you to better understand the differences between different attribution models and visually see the value of each channel at different stages. The models are disassembled in detail in Chapter 4 "Presentation".

Let's take two models for comparison:

by last indirect click (in all standard Google Analytics reports by default);
on the first interaction.

And we'll choose "Change in number of conversions (%)".

As you can see from the comparison table, the data for the two attribution models is very similar. The most significant changes were in the channel (Others, -26.09%). This includes traffic that Analytics was unable to recognize and assign to any other source. For other groups of channels the situation is more or less equal.

Let's take another attribution model for comparison - "Linear", which assigns equal value to all channels in the conversion funnel.

There are already much more significant changes compared to the model "By last indirect click", since the value of each channel in the chain is evenly distributed.

The channel that is most significant compared to the standard Analytics model was found to be "Straight": +9.57% for first interaction and +62.35% for linear attribution model.

You can create your own attribution models in Google Analytics. However, before you do this, take the time to study the 7 standard models built into the system. It's possible that what you want to implement is already available in Analytics.

Multi-channel funnel reporting and attribution assessment are very important when taking a serious and in-depth approach to the formation of advertising budgets and the contribution of each source to the success of the entire project.

Vk.com -

We've released a new book, Social Media Content Marketing: How to Get Inside Your Followers' Heads and Make Them Fall in Love with Your Brand.

Conversion in Yandex Metrica is the ratio of the number of targeted actions to the total number.

More videos on our channel - learn internet marketing with SEMANTICA

Conversion in a broad sense is the ratio of two parameters to each other. For a website, for example, the ratio of those who downloaded the price list to those who came to the site on the title page. The formula will be as follows: 5 people out of 100 who came to the title downloaded the file, 5/100 = 5% (always expressed as a percentage).

Conversion is one of the main indicators of how effectively a website is performing. Depending on the topic of the resource, you can analyze how many visitors bought the product, what type of product, how many people sent the return application form with registration for the event, and so on.

All this is set by creating for each action. Thus, we see that the concepts of conversion and goals are inextricably linked.

A little theory

Goals are predefined actions that a person will take on a website. Customized based on business objectives. For example, they set goals for paying for an order, clicking on the “Add to Cart” button, downloading a presentation, subscribing to a newsletter, and many other options.

The goal is considered achieved if the visitor has fulfilled the condition specified in it.

Visits in which the achievement of a goal is recorded are called target visits.

And finally, conversion is the ratio of the number of targeted visits to the total number of visits.

Goals can be configured for each counter. Those that will be used to analyze direct actions on the site, the size of conversions, and the achievement of goals are called conversion.

How to set up conversions in Metrica

In the Settings section, select “Goals” - “Add goal”

There are 4 types of goals:

Quantity .
Visiting pages.
JavaScript event. Action on a button, submitting a form.
Composite goal.

What conversions to track

First, analyze the website.
Record a list of all useful elements and possible points of contact for visitors. These could be:
-buttons, for example, “Buy in 1 click”, “Order”, “Pay”;
-social network icons are also classified as buttons;
- transition links;
-clickable images;
-forms for feedback, applications, registration, etc.
Write scenarios for the behavior of your visitors on the site. If your target audience (TA) groups are significantly different (suppliers, customers), you need to write down the steps for each. If you do everything right, the key elements will be actively used by your users.
Next, select several options that best match your goals and set goals based on them. For example, use a composite to indicate the steps of the sales funnel.
Wait until the statistics are collected and start analyzing.

Conversion is important to measure and review. For your business, this is double analytics: on the one hand, it will help you create a strategy on how to grow and increase profits, on the other hand, it evaluates how effectively all promotion, optimization, advertising measures - all your marketing - are working now.

How to view website conversion in Metrica

The analytics system provides several tools with which you can monitor the performance of your web resource.

Conversion in metrics reports

To view statistics on tracked conversions in the ready-made “Standard reports”, select “Conversions”.

Click on the target that interests us.

In the window that opens, we receive detailed information in graphical and tabular form. In the settings, specify the desired date interval.

Using Webvisor

This is one of the most interesting and informative services.

In addition to details on goal conversions, it gives an idea of how the visitor behaved on the site, what he did, what he watched, what he couldn’t open. Based on the data from the web viewer, you can safely prepare a document to improve the usability of the site.

In the Goals column you can see an icon with the numbers 1, 2, etc., this indicates that the goal (if 2, then several) was achieved by this visitor during the visit recorded in the service.

Reviewing a significant sample of pre-purchase visits can help understand overall behavioral trends leading to goal completion.

Refinement of standard reports

Yandex Metrica allows you to arrange data quite flexibly, using various slices. In order to evaluate targeted actions, you can add this parameter to standard reports in the same way.

For example, consider the “Sources, summary” report.

Click on the “Metrics” button to open this set of categories.

We are looking for the “Conversions” group. In the first option, we will get a percentage ratio - we will see the share of target visits among all visits.

Depending on what information needs to be displayed in the report, we can also add the “Goal Achievement” group. In this case, a column will appear with the absolute values of the total number of goal achievements among all target visits.

Using the filter settings in the column, you can convert them, if necessary, into percentages.

Additional tool for forms

Yandex Metrica allows you to analyze the completion and conversion of registration, order, and application forms through Form Analytics.

Access to it is located in the general panel on the left in the “Maps” section.

You can select any form from the list (1) and view statistics on it:

conversion;
The “Form Fields” tab displays information about how many people are left after filling out each form field.

Statistics like these often help us take a fresh look at how many fields we offer our visitors to fill out and how difficult they find them.

Is there an Associated Conversions report in Yandex Metrica?

For example, a visitor visited the site several times from different sources: then from advertising, and finally reached the configured goal. Which attribution is chosen affects which traffic source led to the goal. In our case, during attribution, the last transition will be “advertising systems”.

When the system records that a person was on a web resource, and then took a long time to decide on a purchase, and visited several more times from other sources, there is a separate term - deferred conversions in Metrica. Currently, it is at the stage of introducing a tool for “calculating deferred conversion”.

Reports help with the conversion tracking process in Yandex Metrica. By itself, one-time statistics are not very informative. To understand what the dynamics are, you need to compare statistical data for different periods. Ideally, in addition to the counter, you need to maintain internal reporting and analytics on the conversion price in Metrica, accumulating numbers over a long period. Only in this way will all previous efforts to develop a strategy, set up, adjust, bring real measurable results.

Remember, analytics is essentially a ready-made plan of action on how and where to move next; it is worth spending time on.

Offline conversions in Metrica

It is not always the case that customers who place an order through a website pay for it via the Internet. The statistics service allows you to build a connection between online and offline actions and calculate offline conversions.

How does this happen?

Everything is the same, using customized JavaScript event goals.
Set up data import into the analytics system and set goals.
In the settings they set “Extended conversion accounting period”, since payments can be made at intervals of several days after the order is placed.
A prerequisite is the ability to identify site visitors, for example, by discount card number, promotional code, or login.

Today the following systems are integrated with Metrica: Alloka, Callibri, Calltouch, Call-tracking.by, CallTracking.ru, Comagic, Mango Office, PrimeGate, QUON, Ringostat, Roistat, Vector.

As a result, the resulting statistics will help you analyze customer behavior and their path to conversion even better. Plus, it will be possible to target contextual advertising to a new segment of the target audience.