Practical Email [CRM] Marketing

Last month we created a content plan for newsletters .
This time we will return to the topic of analytics in bulk email marketing services the results of campaigns. Again, sent according to the plan need to be measured and somehow interpreted in order to plan your subsequent actions.

You can analyze the email channel at three levels:

We will talk about working with these levels further – as usual, with examples ↓↓

Level 1: Campaign

Campaign -level analytics is probably the most covered in email marketing articles. We prepare a specific newsletter, send it to our subscribers, and then look at what happens.

At this level live such “classic” indicators as:

Here we can also add statistics from web analytics services (Google Analytics, Yandex.Metrica), which we can extract if we mark links in the letter with UTM tags:

Digging deeper, we find a host of additional ways to improve our understanding of campaign results:

For mailings on a specific topic, you can use indicators that characterize their specificity. For example, for letters with product selections in online stores, we pay attention to:

In this way, make sure mobile users find the site it is possible to analyze not only mass mailings sent “manually”, but also individual automatic letters , where we will be interested in the same data.

In general, as you can easily see, there can be a lot of  indicators at the level of individual campaigns. Therefore, sometimes the task arises not to increase their number, but, on the contrary, to decrease it in order to fit the analytics into an acceptable time frame.

One way or another, all this wealth of indicators serves to ensure that we can measure the results of the sent mailing quite accurately and understand a lot about the reaction of subscribers to it.

Here’s what it looks like in practice:

Level 2: Campaign Group (Email Type)

When analyzing the results of a group of campaigns, we are talking about the transition from one mailing to several similar letters , which we can combine according to some feature.

For example, it could be a sequence of mailings announcing some promotion (start / promotion is in full swing / promotion is ending – hurry up). Or a welcome series of automatic letters for new subscribers ( welcome series ).

Here, our focus shifts from individual campaign metrics to aggregate and average metrics:

If we take data from web analytics services, we are interested in the aggregate data on UTM tags used in emails:

We can take a larger scale and talk about the results of not campaigns grouped in some way, but entire types of letters :

The indicators become somewhat smaller, tg data because we take a “larger cell”, but there can still be a dozen or more of them.

For example, here is what the general statistics of automatic emails with product recommendations (abandoned product viewing, abandoned cart, offer of related products, etc.) sent through a specialized service like RetailRocket look like :

Level 3: Channel as a whole

Finally, at the top of the “pyramid” are the results of all campaigns, all groups of letters sent via email on the project – that is, we analyze  the email channel as a whole .

Aggregate indicators for analysis:

We can break down the types of emails to assess how much they contribute to the email channel:

It is also interesting to evaluate the share (%) that the email channel makes up in the overall project indicators received from all traffic sources:

The last metric takes us to the most general level , characterizing the effectiveness of email marketing. If we want to understand how well our mailings work, we only need to look at this indicator. Let’s say it is 10-11% of the project’s turnover: well, we can say that the mailings work at a good, average level.

To compare with other channels, we can use additional metrics such as ROI and  cost per click . As a rule, email has higher ones than other channels (for example, contextual advertising).

Taking into account the specifics of the project and our goals, additional, more “advanced” metrics are also available, such as LTV or how the email channel reduces the percentage of customer churn .

But if we take the case of a small or medium-sized project that does not require in-depth analytics, then the “gentleman’s set” of channel efficiency assessment indicators may look something like this:

Moving between levels

Needless to say, we can move between analytical levels both up and down at any time, clarifying the question that interests us.

We get income from the final two mailings above average by ~300 thousand rubles. The email channel returned to a share of 10% of turnover — the campaign fulfilled its task and stimulated orders.

We noticed the problem at the top level. In addition, refined its location at the middle level, and moved on to solving it at the bottom. In addition, tactical level of individual campaigns. Then we moved back up to evaluate the results achieved.

If our analytics are set up correctly. Again, we measure a sufficient number of indicators (possibly doing this in an automated way . In addition, displaying data in dashboards for mailings), we have the opportunity to freely move from level to level. In addition, gaining more understanding and control over the situation at any given moment in time.