Month: July 2017

Is a negative ROI acceptable when acquiring new customers?



In our last article we talked about the importance of differentiating new from already existing customers. We used the example of a SEA Ad as an illustration. We differentiated new and returning customers and calculated a negative ROI for the newly acquired customers. That is typical and we see that often. In these situations, marketers tend to reduce the bids or cancel the entire campaign. At the first glance, this seems to be the right decision because running profitable campaigns should be every marketer’s core concern.

However, is cancelling the campaign truly the right decision?

Well, not so fast. Halting the campaign might be wrong. To truly decide if a campaign targeted towards new customers really should be stopped or not one must consider the “customer lifetime value”. Customer Lifetime Value accounts for the entire relationship with that customer over time. Drivers for that value are all future revenues and additional contribution margins. Additional future marketing cost to reactivate that customer a detractor of that customer lifetime value. Often the sum of future revenues and margins generated by one customer are bigger than today’s marketing cost. Marketers should consider those future contributions and accept a negative ROI today IF that ROI will turn positive over time.

From one customer to groups of customers – Consider Cohorts

To estimate the “customer lifetime value” marketers should consider groups of customers, so called “cohorts”.  A cohort here is a group of customers with a common characteristic. Typically, this characteristic refers to the campaign with which these customers have been acquired. A cohort also requires a specific time period like a day, calendar week or month. For example marketers should evaluate the lifetime value of all customers that were acquired via Google Adwords within one calendar week (starting point) and observe their additional revenues in the coming weeks.

Below is the example of a group of customers that were acquired in “week 0”, their revenue in that first week and the cumulated net revenue derived from that group of customers over the next 20 weeks:

cumulated revenue

An additional way of observing the behavior of that group is to measure “active customers” and the general cohort’s life. The example below shows that over time about 25% of customers are still “active” (i.e. actively purchasing products and services).

EN Sruvivalanalyse

What’s the result?

Even though a negative ROI has been calculated at the beginning of a campaign, you are now able to observe whether this cohort did generate additional revenues and eventually became profitable over time. In many industries time periods of 3 or 6 months are typical until the breakeven is reached. Industries like dating, gaming and financial products often have to expect breakeven points after a year or even longer.



There most important benefis for cohort and customer lifetime value analysis is an an increased customer reach and stronger growth of the customer base. Campaigns are actively run for longer time periods reaching more customers that all contribute to a campaign’s profitability over time.


The most important requirement for a final evaluation of cohorts is clean customer data and a correct assignment of customers to their cohort. In addition, it takes time (often weeks or even months) until a campaign and its actual ROI can finally be evaluated.

If you want to examine a campaign’s efficiency and the expected ROI shortly after it was launched, you should invest into Customer Lifetime Value PREDICTIONS.

CLV Prediction

A later article will explain the approach towards customer lifetime value prediction in detail.

Differentiating New from Returning Customers


In our previous blog posts we often emphasized the importance of a correct tracking (for example by using UTM parameters) to optimize the profitability of every marketing campaign.
In this article, we want to provide you further advice on how to further optimize your reporting by differentiating between new and returning customers. The objective of this differentiation is again to better analyze and optimize the profitability of all marketing activities.

The Problem

At various Applicata projects we often realize that online marketers optimize their marketing campaigns simply for breakeven or aiming for a slight profit. That is quite alright. But it is this sufficient?

As an example, imagine you run a campaign via Google Adwords which is targeted at the acquisition of new customers. Typically, the campaign’s effectiveness is analyzed regularly with a report that typically looks like this example:

General table

Simply spoken out of 4150 orders and marketing cost of over €100.000 the return on investment (ROI) is 2%. Does this answer the question if you achieved your goal of acquiring new customers? Well – not quite…

Every report should be differentiating new from returning customers

The previous example does not give you any information about your set goal. You do not know whether you did acquire a lot of new customers or if the campaign only attracted existing customers. That is because the report does not differentiate between new and existing customers. If you achieved your goal is still unclear.

Often new customers are a lot more expensive than existing ones, which means that the interpretation of the calculated average values is misleading. We always recommend our customers to precisely differentiate between new and returning customers in every marketing campaign.

Such a report could look like this:

Detailled table about new and returning customers

Now you can see that the average ROI of 2% is made up of the highly positive ROI of 275% of returning customers and the negative ROI of -33% of new customers. In a combined reporting the “cheap” returning customers cross-subsidize the “expensive” new ones enormously.

A marketer should evaluate the report and decide whether the campaign needs to be adapted or not:

  1. Have enough new customers been acquired and has the goal been achieved?
  2. Is the negative ROI of new customers acceptable or should bids and budget be reduced?
  3. Are there other possibilities to reach existing customers? For example, via e-mail or retargeting or push notifications in order to save costs.
  4. Are there special keywords in the campaign which brought back existing customers? Should these keywords maybe be pulled to an extra campaign for returning customers?

Is the reporting of Google Analytics and Adwords enough?


Many companies use the tools of their media partners for their reporting (e.g. Google Adwords Editor, Google Analytics, Facebook Campaign Manager, etc.)

However, you cannot connect these tools to your shop system and as a result cannot differentiate between new and returning customers.

Even Web Analytics tools such as Google Analytics cannot do this. Google Analytics identifies returning traffic but only within 30 days and does not provide any data whether this returning visitors is a new or an existing customer. For example, if a customer did buy something in your online shop a year ago, Google identifies him/her as an “expensive” new customer, although he/she is in all actuality a returning customer.

The Solution

To guarantee a clean and correct reporting of your marketing campaigns we recommend combining data from your tracking solution (e.g. Google Analytics) with your shop data (customers, transactions).

differentiated reporting

In addition to Google Adwords we recommend differentiating between new and returning customers in all your marketing channels (Display, Affiliate, Social Media etc.). The easiest way to do that is to deploy a professional business intelligence solution like Applicata, which also has lots of other advantages.