Choosing The Right Metrics For Paid Marketing

Like life, it’s a balance of quality and speed.

Written with growth marketer, Laurie Bell.

Picking the right metric to evaluate paid performance campaigns is crucial, and not immediately obvious.

It’s a delicate balance to strike.

Balance…get it?

The wrong kind of metric could easily steer you in the wrong direction, either by focusing too much on volume which often comes with low-quality users, or focusing too much on revenue which slows down the process of evaluating your campaigns as you have to wait for the conversion data.

At Heja, we’ve tiptoed towards each direction at times. But like any good startup, we learn our lessons fast and change course for the better.

So it’s experience that tells us what you want to avoid is having to sacrifice the quality or the speed of evaluating the campaign.

The most common ways to analyze paid performance are:

• Customer acquisition cost The advantage of this metric is that it is easy to track and gives you instant feedback on the campaigns. The downside is that you treat all new customers equally. Some will become paying customers and some will never get activated (get into the habit of using the product) and thus churn right away. If you only optimize your campaigns based on CAC, it is very likely that you will get high volume but the low quality of new customers.
• ROAS (return on ad spend) That is all that counts in the end, right? There actually are some disadvantages of choosing revenue as your key metrics for paid marketing. For us at Heja, tracking is a key challenge. After the iOS-14 release, it’s hard to track users from different campaigns inside the app and see how they use in-app purchases. The other main disadvantage is that it slows down your evaluation time, as some new users convert right away whereas some users take months. Paying users also have different value to you, as they all have different churn rates, and some increase or decrease their spend with you over time, etc. Therefore it takes months to attribute a correct value from paying users and evaluate your campaign performance.

This is how we’ve solved these two challenges at Heja

1. We look at activated customers instead of registered customers. A user is activated once it gets into a habit of using a product. This is quality assurance as an activated user is more likely to be retained (keep using the product) with high engagement. At Heja we define activated as a team with at least 6 users. Our data clearly shows that a team with at least 6 users uses the key parts of our product and has high retention.
2. We divide our customers into different cohorts. A cohort in this case is a fixed group of users that behave in a similar way. In our product, we group all users into countries and sports cohorts (Basketball US, Soccer UK, Netball New Zealand, etc.) For cohorts to be meaningful, the users within a specific cohort should have similar behavior. The cohorts should also differ from each other, otherwise, there is no point in creating a different group of users.
3. Calculate the lifetime value (LTV) of an activated customer for the different cohorts. We have created a growth model where we add input values such as conversion rates to a paid subscription, paid subscription retention, average prices, and subscription length. Based on these input values the growth model gives us an estimated LTV for each cohort.
4. As we now know what an activated team is worth to us per cohort, we calculate the LTV for all the activated teams a campaign brings in and compare it to the spend. Let’s say that a campaign has given us 100 new activated basketball teams in the US. If one activated basketball team has an LTV of $50, the estimated cohort value becomes $5,000. If we’ve spent $1,000, the cohort LTV : CAC ratio is 5.

The advantages of this methodology are several. First, we get quick feedback which makes sure we can optimize our campaigns quickly. The activation for our teams is normally is within a week after registration, which enables us to start the evaluation within a week from the campaign start. More importantly, we know exactly how much we are willing to pay for an activated team within one cohort. For a cohort with twice as high LTV, we are willing to pay twice as much. This makes sure that our paid marketing is driving growth for the cohorts with the highest value to us.

To summarize, measuring activated users makes you focus on quality for your paid campaigns.

And because all activated users don’t have the same value, group them into different cohorts and calculate the LTV per cohort.

This enables you to get quick feedback but also makes sure you focus your efforts on the cohorts that bring the most value to you.

That’s how we stay speedy without sacrificing quality.

Balance. 🧘‍♀️

• *Pssst, we’re currently hiring talented people over here.