How to calculate the ROI of an entire project: end-to-end analytics in simple words. End-to-end analytics in Google Analitycs

In this article we will explain in detail what end-to-end analytics is, how to use it, and how to save money on your advertising budget. AND how can you earn money, using end-to-end analytics when setting up the Customer’s advertising campaign. If you are actively advertising in different channels, but cannot calculate the effectiveness of all these activities and find out which gives the most benefit? To do this, you need to set up end-to-end analytics. , guest posts with a link, social networks - now all sources of clients and orders will be taken into account. Read on to find out where to start and how to achieve this result.

a summary report or a single tool that collects statistics on all advertising channels and analyzes conversions;

a detailed picture of where the client came from and what request he used;

a visual path of the client from attraction to the amount of the check that he left with you.

What does end-to-end analytics provide?

CRM, web statistics and CallTracking become one, and you gain insight into which channels perform better.

You will be able to influence the sales funnel.

You'll see which tools don't work so you won't waste time or money on them.

You always know how much an application and a visitor cost.

An end-to-end analytics report allows you to make accurate predictions about the future of your business.

Sales and online statistics will give you an understanding of exactly how your sales mechanism works and what attracts the buyer.

End-to-end analytics saves you time on data aggregation when it is performed by a special tool or specialist.

Peculiarities

The formation of a sales funnel should occur simultaneously with the implementation of end-to-end analytics. It will not be possible to “jump in on the fly” - the reliability of the data may suffer as a result.

End-to-end analytics is useful because at each stage of the buyer’s movement towards the target action there is a different indicator, and from it you can understand whether your ad is effective, conversion percentage, repeat purchases, etc. It is illogical to judge the quality of ads based on completed sales - between There are several intermediate points in these stages, and this may be the case.

It is with the help of end-to-end analytics that you will begin to calculate your ROI indicator and see how well your advertising investments have paid off.

UTM tags must be on all sources in order to distribute all visits and actions clearly across channels.

How to implement?

For end-to-end analytics, you will need a special service like Yandex.Metrica. and how to set it up, read the link. Here, ideally, it will be possible to “link” data from many sources: Yandex.Direct, Google AdWords, Yandex.Market, Facebook, VKontakte, Google Merchant, and also, for example, Podarki.ru. All this can be found out from a specific service before starting work. For example, the list above is Roistat's capabilities, and that's not all.

Roistat

Using end-to-end analytics in Roistat

Offers statistics on 22 indicators and 11 advertising integrations. There is a 14 day free trial. Cost: from 177 to 1623 rubles. per day, depending on the size of your business and the number of projects.

Attention:

The advantages of the service are that you receive a package of useful reports, it has its own good CallTracking, simple integration with advertising channels and analytics for various channels, convenient SMS notifications about expenses and profits, as well as ROI fluctuations. There are also disadvantages: a somewhat abstruse interface, integration with some CRMs occurs with errors, termination of work means receiving data in XLS format, and then as you wish.

Calltouch

End-to-end analytics are also offered by the Calltouch service. Here you can calculate a budget to suit your needs. There is no free trial period in the service, the estimated cost for a small project is 5,400 rubles. To view the interface, you need to register, which is approved by the service manager during business hours. Here you can link Yandex.Metrica and Google Analytics data to get extended SEO data. The service has its own visit statistics, worthy of CallTracking, as well as the ability to order modifications to the tool to suit the personal needs of a specific project. Among the shortcomings: emphasis on calls, the interface requires improvements, there are no ready-made integrations.

Alytics

An end-to-end analytics service with automation of contextual advertising, which it previously specialized in. In addition, it allows you to flexibly integrate with CRM, and do it for free if you top up your account. The approximate cost for three small projects is 5100 rubles.

Using end-to-end analytics in Calltouch

There are some peculiarities: there are few advertising channels for automatic tracking, visits are displayed only from Google Analytics data (there is nothing to compare with). Among other things, the service is quite young, and not all processes have been streamlined yet.

Google Analytics

You can combine everything in Google Analytics, but you need special knowledge or someone who will control the entire array of data, set parameters and draw the right conclusions. This is either through an acquaintance or for a fee. The possibilities are quite wide, so it’s definitely worth paying attention to this option, even for the money. A few more arguments in favor of delegation:

working with formulas is difficult;

it’s hard to “make friends” with the interface if you are a beginner, and the ultimate goal is end-to-end analytics;

Without call tracking nothing will happen.

As a conclusion: Google Analytics is needed by large projects or experienced perfectionists.

Who doesn't need it?

There is a specificity of business for which end-to-end analytics is not urgently needed. If you sell cars or apartments, there is a big risk of losing a client. End-to-end analytics is designed for a slightly shorter decision-making cycle and evaluation of the results of advertising events. In this case, the costs of end-to-end analytics may be unreasonably high. If you sell high-margin goods, and business is objectively going well, you are looking for a new sales market, don’t get hung up on end-to-end analytics - you don’t need it. If your product is extremely popular in a short period of time, and the fashion for it quickly passes, quickly create landing pages, “pour” contextual advertising on them, and don’t think about complex analytics. A regular application report will be enough for you.

Today we will talk about end-to-end analytics and how to implement this tool inside Google Analytics.

What are we going to talk about:

Why do you need end-to-end analytics?

How to implement it in practice?

What errors can there be and how to solve them?

What is end-to-end analytics?

What situation might you have - a large number of different advertising channels through which users find your site. These may be channels that we manage, such as paid advertising, contextual advertising or special communication channels. This could be advertising that we manage indirectly - for example, search engine marketing in search engine results. These can be any advertising channels.

Simply put, we invest some money, for example 1000 USD, in our marketing. These 1000 dollars are distributed in some shares to different communication channels:

$300 for search engine optimization,

$100 for price platforms.

That is, we can distribute these amounts evenly or unevenly. And we need to understand how effective our investments are, and which ones work out better. For example, where is it better to invest - in SEO, Yandex Direct or Google Adwords. Or close everything and work only with price platforms or some aggregators. Only after analysis will we understand when there will be a greater return on the money we invest.

How to calculate the return and efficiency of these investments in different advertising channels?

At the moment, most businesses or marketers who work with Google Analytics consider cost-effectiveness solely based on some kind of user communication actions. Those. the user can call or leave a request, or even, if this is the case of e-commerce, go through the entire ordering procedure and at the last step click “Place an order”. It's great if we have the ability to accept payment from the user at this step or at one of the last steps. When the user has the option to enter their credit card information or pay via PayPal. In this case, we clearly understand that these are the people who placed the order, these are the people who bought from us.

This is ideal. But in reality? The number of transactions on the Internet using cards or other payment instruments is relatively small. I have statistics for last year or the year before, something in the region of 3-5%, this is not even a dominant percentage, i.e. it's not 50%. We understand that a large number of audiences pay whatever they want.

And the fact that the user communicated (filled out an application to purchase a product or sent us a letter or used an online chat) does not tell us about the fact of the purchase. It is rather a certain desire to buy, a certain step before a real purchase. And if we look carefully, and for your real businesses we take the number of these communications (i.e. the number of times where each individual user contacted the business with the desire to buy) and the number of purchases we will see: - the number of communications we have is much greater than the number shopping.

For example, 100 users went through the entire checkout funnel and only 50 of them actually paid.

Users may not pay for various reasons - they may not like something at the last moment, there may be a million situations why the user did not buy, but, nevertheless, such a situation arises quite often.

What is displayed in Google Analytics?

In Google Analytics we display (if the goals or e-commerce block are configured correctly) information about transactions, products that the user has selected. In this case, we see, sort of, sales. But these sales amounts or the number of transactions that we have inside Google Analytics may not coincide with what we have in CRM. If so, then we may actually make incorrect decisions regarding the effectiveness of these channels.

For example, we see that we have a lot of people leaving a request or filling out a free search basket from Yandex, Google, but at the same time the number of people who paid is much less. In this case, when we calculate the effectiveness of this channel, if we only take into account those actions, only applications and not actual fees, we may be greatly mistaken in the effectiveness of this particular traffic source.

Maybe we, say another Google Adwords source, have a smaller number of completed purchase requests, but at the same time the conversion into real payment there is quite high (100 people go completely to the shopping cart with Google Adwords, 95 of them pay). And we understand that in this case this channel will be clearly more effective for us than a free Google search if we work with the tool

The second point is the moment of taking into account all communications. Users can contact us in a variety of ways. We can quite easily and simply track the fact of communication, i.e. what happens directly on our website. If a user left or filled out a feedback form or went through the entire checkout process, we will definitely see that this person did something useful for us.

But if the user called us, sent an email to our regular mailbox, if he came to the store, if he used online chat, then by default all this information is not collected. And for a very large number of businesses, we may have a situation where the number of these communications via call, letter, offline and online chat will be more than 50-60-70%, and the number of people who fill out the form may not be very large, but in fact, we only record them, only those people who filled out the form with us.

Let us formulate three big problems:

1. We do not see real payments from those linked to advertising or non-advertising channels that we use.

2. We do not take into account all communications by default. Those. Thus, we cannot compare the data that we have in CRM - that this particular order came from this particular traffic source and the user used, say, a call, a letter or something else. If so, it is quite difficult to make decisions about the effectiveness of advertising channels inside Google Analytics. And then how to make decisions about the effectiveness of a particular channel or advertising campaign within Yandex Direct. and in general any entities that we transfer to Google Analytics (regions, social media, any other information) are difficult. We actually don't know any of this. In fact, the data we have inside Google Analytics cannot be directly trusted.

3. We do not take into account returns

According to the law, in different countries, within 14 days, the user can return the product without explanation if it remains in salable condition. If you have a business related to clothing, if you have a business related to some things that may potentially not suit the user, then the number of returns may be large. And by default they are also not displayed unless this is implemented additionally.

1. the user did not understand how he contacted us, or rather, we understand, but we cannot connect this with the advertising channel

2. we do not see the exact payment

3. We do not accept returns

The analytics process works as follows - we have some kind of advertising, this advertising generates some kind of traffic, users who visit the site, and actually some of these users become leads and they go to the sales department.

But in fact, in order to correctly collect data and correctly take into account the effectiveness of traffic sources, we need a different scheme. We have traffic that is generated by visitors. These visitors can use a variety of tools to communicate with businesses, and We must take each of these tools into account within CRM. But at the same time, we must also understand within CRM that this user who used the online chat came from a specific advertising campaign, for example, Google Adwords for some key request. And as soon as a decision is made here in CRM, i.e. in fact, the manager who is handling this order will set the status that this client has paid, we immediately transfer information about this successful payment to the analytics system. We can also transmit all sorts of communication things in order to build funnels and see the effectiveness of not only advertising tools, but also other methods of communication.

What does the new analytics system look like in practice?

We may have such a situation - we have implemented a callbackhunter, which gives us a certain number of leads and applications. And we think - oh, how cool, great, a good tool, we’ll pay, but if we see the overall picture of the redistribution of ways to contact us, we will see that this callbackhunter can simply take a bite out of other channels - phones, forms, online chat

Thus, before the implementation of callbackhunter we had 100 calls per month and after the implementation of callbackhunter the same 100 calls per month. Those. it did not bring any additional value to the business

Accordingly, from the site, by default, data should also be transferred to analytics systems, but here there is more information about how the user interacted with the site itself, how he looked at various sections, how he looked at the product card, how he clicked on something important to us buttons and so on.

And after we take and implement such a scheme

If we take into account each communication channel, if we take into account returns, if we only transfer data from CRM to Google Analytics from the moment of actual payment by the user (only at this moment) - only then can we build such beautiful funnels. And then we will have a complete picture of the effectiveness of our marketing. We may have different advertising channels that turn into applications, i.e. users make some kind of communication, then at this stage we see the situation regarding the effectiveness of advertising channels in terms of applications. We see how much one application from Yandex Direct, or one application from VKontakte, or one application from somewhere else costs us. We see the number of applications, we see the conversion rate (i.e. the ratio of those people who left an application to the traffic that came to the site in general).

Actually, we can see deeper, we can see the deal. This is not just an application, but let’s say a transaction, where some kind of communication between the manager and the user has already taken place, they draw up some details, i.e. not just an application, but a transaction, and, accordingly, we can also transfer this information to Google Analytics.

We also have information about paid transactions, i.e. when the user actually bought. And then, here, we can see the effectiveness of the selling price. Efficiency in terms of ROI, efficiency in terms of profit for each of the advertising channels, campaigns and any other entity.

For example, we can clearly understand that Yandex Direct brought us 100 leads, but there were 2 sales, and the ROI, let’s say, is 103%. But Google Adwords brought us 50 leads, less than Yandex Direct, but at the same time we had not 2 sales, but 48 sales. Let's say the conversion from applications to sales is much higher with Google Adwords and then our ROI is 500%. Conclusion: despite the fact that Yandex Direct brings us more applications for us. but from the point of view of business, money, profit, Google Adwords is effective.

But without such a funnel we will never say this, if we do not set up all the nuances in the analytics, then we will be blind.

It is very important to see not only the situation associated with one sale. There are many businesses where we can have much more sales per user than one, and even more often a situation arises where our marketing may pay off not from the first sale, but from the second, the third, the fifth.

It is necessary to take into account all the sales that this specific user made, who came through this specific advertising channel Yandex Direct, Google Adwords or whatever. In this case, if we implement this scheme, we will have a complete picture of how to implement this matter.

We solve problems of end-to-end analytics

There are many different services that solve this problem automatically. Those. some kind of customization is needed, some of the services are more focused on CRM, some are more focused on managing contextual advertising, some are more focused on call tracking, some are purely end-to-end analytics systems. I gave three examples, but there are more. We can implement an end-to-end analytics system using these services, and they are more or less tailored to solve the problems that we face so that we can see the entire funnel, but these services have many different disadvantages, each with its own.

Probably the main disadvantage of all these services is that they are fairly closed systems tailored to one narrow task, either high-quality tracking of calls and end-to-end analytics, or managing contextual advertising and somewhere end-to-end analytics. And the downside is that they are difficult to modify if we have new tasks. Therefore, today we are talking about how to implement this task using Google Analytics, namely, so that within our analytics we can see in full a report like this, where we have an advertising source, any - paid or free, where we have a completely information on the stage of interaction, our costs, the effectiveness of our ads, where we have information on profitability.

We can enter communication goals into such tables, and, in fact, inside Google Analytics we can measure each of the stages with metrics that best show us the effectiveness of a particular step. We can have several reports, each report will show different metrics. Then we will see the big picture, in fact, see our entire funnel.

How to set up Google Analytics correctly?

First, you need to use advanced e-commerce functionality.

The advantage of expanded e-commerce is that we can take not everything that is implemented there, but only some part, namely the part about specific payment, and actually not take all the other functionality. That is, implement it in parts. But in this case, our implementation scheme will change. And at the last step, on the thank you page we will not have the same code that transmits information about the transaction to Google Analytics, but will be the penultimate step of the funnel. Roughly speaking, at the last step (transaction) on our website, we will simply transmit information that the user has completed this step, that he has somehow communicated with us.

Moreover, we can transmit this information using such functionality as check out options, not only for forms on the site, some ordering procedures other than check out, but also for any other way for the user to contact us. We can see that the user went to check out or called - and for us this will be equivalent. But just the last step - we will transfer information about the transaction from the CRM.

In order to implement this entire system we need three steps:

1) remember the user id from any communication channel. Those. it doesn’t matter how the user contacted us - he can call, he can write to us by email, he can go offline or do anything. We need to record its unique identifier from each channel. I will say a little more - this is a unique identifier that Google Analytics gives to each user.

2) The second thing we need to do is upload expenses for advertising tools. This is probably a less mandatory step, but it is important from the point of view of calculating various indicators such as ROI and so on. Those. Without this everything will work, but with this we will have much more data.

3) The third step - we transfer successful payments from CRM to Google Analytics only upon actual payment. Moreover, we transmit both payments (the fact of payments) and returns (if we have them).

Setup. Step one

If you use your Chrome browser, let’s say, go to any site where Google Analytics is installed (in general, this is almost any site), not counting some social networks, right-click, click “view code”, then you will have an additional block, which can be on the right or below.

Now this is called application, previously resources, where we have such a tab and a Cookies sub-item, if we go there, we will see a cookie called GA and this unique number is clientID.

Those. this is a unique identifier that is issued to the user (each browser) by Google Analytics. If a user visits, for example, the same site a second time, then his unique identifier will be exactly the same. But if he comes to the site for the first time, this identifier will be different.

Our task:

— take this clientID. Using JavaScript, PHP or any technology we can get this clientID.

- we can transmit it along with some kind of communication. For example, if this is a procedure for placing orders through the website. Here we have standard fields: first name, last name, phone number, what the user wants to buy, etc., i.e. a lot of different information. We can add another field, which we will call clientID and, taking this unique identifier, transfer it to our CRM, thus expanding it by one additional field.

If the user called us, we can also take and get this identifier. clientID is a unique identifier to which we associate all the information within Google Analytics about this user’s visits. For example, the initial traffic source through which it came, etc.

If we see this clientID in our CRM, we can then transfer it and expand Google Analytics’ knowledge of what exactly this user bought. Accordingly, we have different communication channels - call, letter, online form, online chat. Each of these communication channels can be taken into account.

Call- here we need dynamic call tracking, with which we can replace the phone number for each client, user, who is recorded in Google Analytics. Accordingly, upon a user’s call, we will be able to obtain a unique clientID from the call tracking system.

Letter we can send to the address [email protected]. The letter will be sent to [email protected], but at the same time we will retain our identifier. Those. We will show our unique mailing address for each user, but all mail will be sent to our main mailing address [email protected].

If this form, then everything is very simple here. If a user places an order with us through the website or submits a form, then along with his contact information we also take the clientID, a unique identifier, and transmit it.

If this offline, then we can display a banner on the site with a QR code in which this clientID will be encoded and oblige our managers to ask for the presence of this QR code; if it is, then we read it with a regular mobile phone with a camera. Thus, this code will be saved with us.

If this online chat, then here we have the opportunity to transfer some custom field to the administrative panel, which the sales manager works with - there will be a name, correspondence, phone number and, among other things, clientID.

Setup. Step two

There are many services that allow you to do this. Most of them are paid. For example, OWOX BI - they have the ability to import data on advertising costs.

Setup. Step three

The third step (the most important) is transferring data from CRM to Google Analytics. This step is implemented using Measure Protocol technology. This is one of the APIs that Google Analytics has internally. With this API we can expand the data we have. Those. data, on the one hand, is collected from us using regular code that we place on the site. And another way is to transfer data from any place where we have Internet. This could be a CRM, any website or device connected to the Internet. From there we can transfer some information using the Measure Protocol to Google Analytics. In particular, in solving our problem, we can transmit information about the successful transaction that the user made.

The last step. Analyzing ready data

In this case, we receive ready-made data - we have a source, channels, sessions, costs, profit, price per transaction, number of transactions, number of conversions and ROI (already ROAS). We can add some other necessary and important data to such reports, but, in any case, here we can fully see the situation that we have. And then we can clearly say which advertising channel is more effective or less effective for us.

We can visualize our reports. For example, we can use Google DataStudio to display all this information visually and work with this information and conveniently analyze advertising campaigns.

When transferring data via measure protocol, a new session is created within Google Analytics for this user and this session with direct/none.

Problems arise here from time to time, because the attribution model that exists inside Google Analytics is the last indirect click, and if the user has already visited once through direct, if we have a long enough sales procedure, he could come back to the site after the initial advertising channel and , accordingly, the last indirect click will not be our initial source through which he made the conversion, but, say, another channel or the same direct.

What can we do?

We can overcome this in the following way - we can, along with clientID, transmit the value of this metric if this is a paid advertising channel, or the value of the source and channel of traffic that the user used when he made the communication.

Attribution only works for channel sources, not regions or other session variables.

If we need such additional characteristics, for example, where the user came to us from and what he bought, from the point of view of the city, then we can also transfer this information to CRM, we will link this data, and then our transaction will be recorded exactly to the region where the user was when he left the request.

Measure protocol may not always work correctly and this must be controlled. Periodically, problems arise with transmitting data over this protocol if we use POST requests; often there are fewer problems if we use GET requests. Experience suggests that when implementing this scheme, it is advisable to have a log file in which to record each request that was sent by Google Analytics, because Google Analytics may simply respond that everything is fine, but in reality everything may not be fine. If we have a log file we can quickly find any errors.

If the CRM is a mess, then Google Analytics will not help.

For example, if we have currencies inside CRM written differently - somewhere we have rubles, somewhere dollars, somewhere hryvnia, somewhere euros. And, if we do not bring this data into one form or do not take it into account at all when implementing the entire scheme, then our Google Analytics data will be inaccurate.

The same applies to the payment accounting process itself - if we have many different statuses that we have or may have upon successful payment, then here we also need to be very careful when implementing this entire scheme, so that there are no moments when the payment is fixed, but in fact, the money has not come and will not come.

The topic of end-to-end analytics has recently been reminiscent of dark matter: everyone knows about it, heard it many times, some even have an idea what it is, but when you try to find out the details, they go into abstract reasoning about the creation of the universe and user ID.

In fact, there is nothing complicated in end-to-end analytics - this is the most common analytics that tracks the entire user journey from the first contact (be it a search query or a call on a business card) to repeat sales. It is thanks to this “permeation” of the entire process that the name “end-to-end” was established. But, hand on heart, I believe that any analytics must be end-to-end, otherwise it is not analytics, but assumptions in a specific area of marketing interaction.

Why is not all analytics end-to-end?

So, imagine the situation: you have a small online store, you run ads only in Google.Ads (formerly AdWords), the site has Google.Analytics configured with advanced (this is important) e-commerce. All sales are made exclusively through the website, payment is made online (including returns). In this case, the functionality of Google.Analytics is sufficient for the work. It is the very instrument of end-to-end analytics, since it records visits to the site, the process of selecting a product, how the client put something in the cart, deleted it from it, what promotional code was used. If after a month this user returns and makes a second purchase, this can also be caught using standard Google methods.

But, alas, in practice everything is a little different:

— stores use many channels to attract users;
— offline advertising is often involved;
— communications take place not only on the website, but also by phone (sometimes clients come to the office!);
— not all “requests” (or leads) are sales;
- even if the sale has taken place, the client can return the goods within 14 days (or later, in accordance with the contract).

As a result, it is almost impossible to calculate ROI for each individual channel. But in practice, alas, people don’t even combine expenses from different sources, let alone combine data from different stages and processes. As a result, they receive scattered information that is in no way connected with each other:

Therefore, wrong decisions are often made based on myths, rumors and intuition (which is based on myths and rumors).

What to do about it?

And so, when it seemed that everything was lost, end-to-end analytics came to the rescue, which with high accuracy (down to each individual user) helped to connect disparate data into a single chain.

There are three main approaches to solving the problem.

We summarize all the data in Google.Analytics

The first approach is as follows: we bring everything together in Google.Analytics, transferring all subsequent stages in various ways. For example, almost all CallTracking systems, hunters and online chats send standard events to Google, which you can set a goal for and attach a value to. CRM systems can also transfer the necessary parameters if configured correctly. All this is linked by user ID. As a result, we can track every single customer.

Similarly, we add expense data to Google.Analytics (for example, about ).

As a result, all points of contact are recorded, and sales are carried out using standard means of expanded E-commerce.

Using specialized services

The second method is less inventive and much simpler: you buy a subscription from a specialized service (for example, Roistat or Alytics), which brings all the data together in their interface.

There are no difficulties with settings and connection. The system also contains the most popular standard reports and sales funnels; you don’t even need to come up with your own.

Self-summarization of all data in tables

The third method is the most resource-intensive at first, but simple and convenient in the future. Upload all data from different sources into a single database (this can be any cloud solution, your own database on the server, or regular Excel/Google.Sheets tables) and build any reports that are needed (for example, using Google Data Studio or Power BI).

The advantages and disadvantages of each method are summarized in the table:

Data reduction method	pros	Minuses
Aggregation in Google.Analytics	- for free; — does not require additional integrations; - no need to additionally learn something new.	— limited to the functions and reports of Analytics itself; — some data may be lost.
Third-party end-to-end analytics systems	- fast; — there are ready-made reports; - official technical support.	- you need to pay every month.
Summary in tables	- for free; — you can build any report for any period.	— quite resource-intensive to set up; — you need to be able to build queries to the database; - you need to be able to use analysis systems.

What do we get in the end?

As a result of using one of these methods, we can get the following result:

I would like to note right away that in this case we used a simplified calculation model; fixed costs were evenly distributed among all sources.

Separately, I would like to draw attention to multi-channel sales. I recommend using a linear model if income is evenly distributed among all channels involved in the sale.

In this case, we immediately see the results from each specific channel. Imagine all this in dynamics... (I would like to note that “Recommendations” is the result of advertising costs in previous months. You should not assume that it is completely free. In this case, again, they wanted to show the overall picture).

What do you need to get started?

To start using all the benefits of end-to-end analytics, you need to be prepared both mentally and technically. The fact is that, firstly, you will have a fairly large amount of new data to analyze; this scares many. Secondly, from experience I can say that many people do not even do such simple things as UTM tags, which makes further analysis impossible.

Please note that if AdWords accounts are linked to Google.Analytics, and Yandex.Direct is linked to Metrica, the data is transferred without labels, automatically (of course, there are labels - Google and Yandex set them themselves, but their own, specific ones). Third-party systems (call tracking and CRM) cannot read these tags, so you must use standard UTM.

Also, ensure that all systems read and write this data correctly. Only then will it be possible to link them together.

Instead of a conclusion

Regardless of how you analyze the data, remember that analytics is not needed for its own sake, but for making decisions. Even the most beautiful dashboard is just a picture if it doesn’t answer the question you asked.

It should be remembered that analytics is not an add-on to the sales process, but an integral part of each link in the chain:

And these are not all different analysts, but one, connecting chain.

And building a report based on the collected data is not so difficult:

In one of the following articles, we will look at a specific case for setting up end-to-end analytics and show the results of its implementation.

Pump up your SEO skills to the maximum! Author's courses SEO-Koksharov (Devaka)

Advanced course:
On October 17, the Hard SEO course “From Specialist to Pro” will begin.
Course duration: 6 weeks.
You will learn in-depth website analysis, understand search engine algorithms, and use advanced SEO tools.

Course for project promotion:
October 22 – the author’s SEO Pro course, created in collaboration with WebPromoExperts.
Course duration: 4 weeks.
You will learn how to conduct an SEO audit of websites, analyze semantics, increase website link mass, and analyze the effectiveness of search promotion.

What does every business owner want? Know exactly the amount of advertising expenses and understand how much money you managed to earn from each advertising method. Standard analytical methods (clicks, impressions, transitions) do not provide the necessary amount of information with which you can determine effective and ineffective channels for attracting customers. These problems are solved by end-to-end analytics.

Order website promotion

Ingate in the new issue
"Transformer"!

Watch the video and learn digital tips from the company’s top officials.
Find clients. Faster!

Three levels of analytics

Given the analytics frameworks described above, three levels of complexity can be defined.

First: everything by hand. A suitable option for small and micro businesses that cannot afford paid analytical solutions. It requires the following set of tools:

Spreadsheet Google, Excel;
electronic platform for receiving applications (CRM);
Google Analytics, Yandex.Metrica. You can use the services in a complex;
call tracking system (automatic or semi-automatic).

Action scheme: the analytics system records user actions on the site. When the application is submitted, information about the source enters the CRM. In the elementary model, the tools are not connected, so the data must be uploaded manually, combining them according to the required parameters. This is convenient for low traffic and budget. As soon as the company overcomes this stage, it needs to choose a more complex solution.

Second: automation. A solution for small or medium-sized businesses with a large volume of operations. Suitable Tools:

Google Analytics;
automatic call tracking system;
tool for automatic import of expenses;
if necessary - Microsoft Power BI.

The implementation period takes on average 7 days. The principle is based on setting up a connection between tools for automated data upload to Google Analytics. Most modern systems support this feature.

Despite all the advantages of automation, this solution has its own characteristics. There is no access to each user within one session. If the number of sessions exceeds 40 thousand per day, the system can sample data. In addition, the system prohibits the transfer of personal information.

Third: perfect analytics. This level is chosen by representatives of medium and large businesses. It allows you to automate processes without the problems mentioned above. Medium Tool Set:

Google Analytics;
automatic call tracking system;
database (preferably cloud);
Excel or more advanced tools (Tableau, QlikView).

The principle can be described briefly: all received analytical data automatically enters a single database in its pure form without distortion. Thanks to this, you can obtain a sample for any request: study the behavior of each specific user, analyze groups of buyers, test a huge number of marketing hypotheses. There are no restrictions. The obtained data can be downloaded in the form of convenient tables and graphs.

Three stages of setting up end-to-end analytics

When implementing a system, the services of programmers and integrators are required. When everything is adjusted and tested, you can use the tools without involving IT specialists. It is better to implement end-to-end analytics when building a sales funnel. Without this, it will not be possible to estimate advertising budgets. In general, all work can be divided into three stages.

01.

Selection of key control points

The number of business indicators available for tracking can be unlimited. At each stage, these indicators may change. You can select those that are relevant for a given time period. For example, key indicators during the period of attracting customers may be clicks on advertisements, during the stage of sales growth - average bill and conversion, and during retention - repeat orders.

Among the most important indicators you need to take into account ROI - return on investment. This is the amount of money that was returned by spending money on advertising. In essence, end-to-end analytics comes down to calculating how much the company earned through advertising and how much each channel pays off.

02.

Setting up analytics

Integration of used services into a unified analytics system. Setting up UTM tags and generating a link that can be used to track the sources of visitors. Customize link markup to get accurate information for each channel.

03.

Metrics tracking

Recruiting a mass of events and visitors, tracking channels using UTM tags, using analytics and call tracking systems to analyze target actions. All indicators are fed into a unified analytics system, where there are tools for their convenient calculation and forecasting.

Market leaders choose Ingate

How to become an Ingate client?

To order a strategy in Moscow or another city, simply fill out an application. A customer service manager will contact you and prepare a personal commercial offer from the performance-marketing agency Ingate, tailored to your goals and capabilities.

Reading time: 7 minutes

End-to-end analytics is a tool that allows you to evaluate the effectiveness of your advertising. It gives a clear understanding of the financial result each ruble spent brings. We’ll look at how to implement end-to-end analytics, set it up and get the maximum benefit from it in this material.

Ways to build end-to-end analytics

Before implementing end-to-end analytics, you need to determine the main goal: what exactly you want to track and what metrics to calculate. Depending on the tools used, three methods can be distinguished.

Whatever method you choose, end-to-end analytics is impossible without combining information from different sources. Advertising data is provided by advertising systems: Yandex.Direct and Google Adwords. Sales information is available in CRM. Call statistics are collected using dynamic call tracking.

After combining all this data, a general picture emerges about the effectiveness of advertising. Let's take a closer look at how end-to-end analytics is built using the CoMagic service.

Combining data in CoMagic

The CoMagic service collects all marketing and sales data in your personal account. As a result, end-to-end analytics reports become available, allowing you to understand which advertising channels are effective in terms of the number of calls, cost of call, and return on investment (ROI).

Where does CoMagic get its data?

In the General settings of the site, in the Integration with services tab, select Advertising systems and enter the account identifier to which the advertised site is linked.

If you have a phone number on your website and customers often call you, your end-to-end analytics will be incomplete without call data. To see all calls from site visitors, we connect call tracking.

When dynamic call tracking is enabled, each visitor is shown a unique number. It is assigned to him for a certain time, which allows you to analyze the behavior of this visitor: from the key request for which he came to the completion of the transaction.

Suppose there is a keyword “buy a car”, you see how many visits it brought, you see how much a call for this word cost, but was it targeted? Did this keyword lead to a deal or not? What is its sales conversion rate? If they not only called using this keyword, but also made a purchase, then you can accurately evaluate - yes, this word works.

In order to see not only visits and requests to the company, but also the transactions themselves, sales data is needed. Then you will be able to analyze not only “how many visits this keyword brought me,” but also specifically “how much money I earned from it.”

Sales - pulling from CRM

The next step is to combine marketing data with sales data - we integrate your CRM system and CoMagic.

Most integrations are done in just a few clicks, allowing you to do them yourself. We have developed integrations with the most common CRM systems. But CoMagic is able to collect data from any CRM, even a self-written one. In this case, integration is performed using the API.

Any integration is free and carries only advantages - additional data for your analytics.

After connecting CRM, the List of Deals report appears in your personal account.

This report can be configured differently, depending on what exactly you want to track. What you can see here: date, type of request, from which advertising campaign, which keyword worked, transaction amount, visitor ID (using it you can go into the visitor’s card and see the entire history of interaction with him), etc.

You can now add additional columns to the End-to-End Analytics report – Financial data. This financial data is the very return on advertising! How many sales were there, what was the conversion, average check, total revenue, ROI. You need to click the Customize Columns button and select the metrics you want.

You can select all columns or customize end-to-end analytics. We recommend paying attention to indicators such as expenses, visits, requests, total sales, revenue, ROI, average bill. The obtained data can be downloaded as a ready-made report in PDF, Excel, CSV or Google Sheets.

But if you are a convinced visual person, speed and ease of perception are important to you, or you need to show your boss beautiful and understandable graphs, and not a pile of numbers, use dashboards.

These are a kind of showcase of your key indicators. Customize your graphs by setting thresholds and always have all the metrics you need at your fingertips in a simple and understandable way.

Dashboards work in Real Time mode, which will allow you to always be aware: advertising is working perfectly, increasing traffic and calls, or the lead plan is going to hell and urgent action needs to be taken.

The resulting visual report can be downloaded in PDF or PNG format. Or set up sending dashboards by email.

Usage

So, we have set up end-to-end analytics and connected dashboards for convenient information analysis. What's next?

End-to-end analytics only works and brings real value when you not only observe, but also make business decisions based on it. After 1-2 months of advertising and tracking all indicators, you can adequately analyze what works in your advertising and what doesn’t.

For example, you have a set of keywords. From month to month, only 2 or 3 of them bring real deals. For the rest, there are no requests or no transactions are made. Increase bids on ads for working keywords, thereby increasing conversion. Having all the necessary data, you can fully optimize your advertising and thereby achieve greater effectiveness.

Analyze transitions across all advertising channels, calculate ROI, try new ways to attract traffic to the site. End-to-end analytics makes it possible to see all your advertising at a glance: what works and what doesn’t. Use this vision to make your advertising even more effective and your business successful.

Are you an expert in
internet marketing?

Publish material on our blog