Growth Marketing Measurement - From Zero to Hero

I frequently get questions about how to approach Growth Marketing measurements - questions range from the latest tools and techniques to addressing the not-so-new privacy changes and SkAdNetwork (introduced with iOS 14).

My response is consistently nuanced—it hinges on several factors, which I'll delve into below. Based on my experiences at companies like Uber and Twitter, as well as various ventures, I'm writing this blog to provide insights that can guide founders, growth marketing teams, and data teams to navigate this topic.

In my view, four critical factors must be considered when approaching the topic of measurement.

TL;DR

1 - The stage of maturity of your company

This is the first one and probably the most important one to take into consideration, and it will be vastly different if you're just starting your business and you're validating or testing your acquisition strategy (most likely to get a baseline for your CAC metric) versus you want to optimize your marketing $ across multiple existing channels.

A - Pre-seed / seed

At this stage, you're most likely trying to figure out your first acquisition channel, dipping your toe in the water to have a baseline for CAC and volume to validate your business plan assumptions (hopefully you have some 😄).

In that case, you very much want to keep your measurement approach as simple as possible since there shouldn't be too much noise in your data simply put, since you only have 1 or 2 acquisition channels, you don't need to isolate the impact from other sources.

From a tooling perspective, you should keep it rather simple, especially if you have scarce resources since sophistication won't give you critical information you can't get:

• Use platform (self-reported) metrics from your advertising channels (likely Meta, Google, TikTok and/or Apple Search Ads) to report and optimize your key metrics (see next sections). It should be your main source of truth at this stage to understand your CAC.
• Use Google Analytics, Play Store, and App Store data to gain more intuition around your funnel efficiency but don't try to attribute your marketing results based on those tools (yet).
• While you want to keep your 3rd party measurement stack simple, it's, however, very important to start designing and building your 1st party tracking metrics (conversion events, key behaviour, funnel...) to ensure you have a good grasp of your full acquisition funnel and more importantly understand how the users you acquire from marketing behave and if they behave differently from your first adopters (Check my post on cohort analysis).

B - Early stage(PMF)/Growth

At this stage, you hopefully have 1 or 2 good and steady acquisition channels, and you have a good grasp of your CAC and acquisition metrics. You've either already established your product market fit (PMF) or you're getting closer to meeting it.

The main questions and challenges usually are "How do I scale my acquisition engine (doubling down on existing channels, adding new channels...) while keeping the unit economics stable" and "Are my existing channels delivering the incremental value I think they do or should I improve my attribution model". Another common question is "How can I track down funnel metrics (beyond acquisition) at a channel level since I don't get user-level data for all my users".

If this is the stage you are in, then you want to start building a more robust attribution strategy that can help you with those questions. Chances are you might also have a hard time tracking your organic channels and understanding where exactly your organic (or usually untracked) traffic comes from. While organic traffic cannibalization shouldn't be too much of a concern at this stage, it's still worth preparing your measurement capabilities to be able to address this question in the future.

This is when you need to focus on getting a single source of truth for your acquisition channels and as much as possible apply the same logic (attribution window, click vs. view-through...) to all your channels to ensure you compare apples to apples.

When it comes to tools, you must start equipping your team with better tools:

• Implement a Mobile Measurement Platform and Google Analytics to build your single source of truth and attribution model.
• Stop using platform self-reported data for attribution but keep using them for daily/granular optimization and specific experiments.
• When it comes to your 1st party data, you hopefully have a data team that can help you with ingesting your MMP, Google Analytics, and spend data and stitch it with your internal metrics.

Regarding user-level information, there are 2 ways to go about it.

• You focus on channels where getting user-level information is achievable, such as affiliate channels and start optimizing those channels based on that data (usually first weeks retention or CLV).
• Leveraging your MMP data, you take the fraction of users that you can track (have opted-in for iOS tracking) and extrapolate to all users coming from this channel.

C - Maturity

At this stage, it's all about optimization of your marketing budget and incrementality. You have a robust attribution model and a single source of truth.

The common challenges at this stage you want to solve are "How do I allocate my marketing $ in the best possible way and as frequently as possible (weekly, monthly, quarterly...)" and "Should my company invest additional budget on Growth Marketing, what would be the incremental value".

If you're at this stage, you probably have a good idea of how to address those challenges or at least where you want your measurement capabilities to be.

Your main focus should be:

• Measure marginal CAC
• Develop a predictive LTV model
• Measure incremental lift and/or causal impact of your main channels
• Invest in a Media Mix Model to understand the effectiveness of your channel mix, address data limitations and have a robust budget planning/allocation system.

As far as tools are concerned, there are a lot of MMM solutions on the market starting with open-source solutions like Robyn developed by Meta.

The most important lever at this stage is Data Science capabilities which you mostly have to get internally.

2 - Your vertical and business model

The second factor is your vertical and your business model. Whether you are a web-only, app-only, or mixed-platform service will impact your approach to measurement and your choice of tools.

• App only, then MMP would be your go-to solution.
• Web-only, Google Analytics would be your first choice.
• Mix platform, you'll have to deal with both solutions and think cross-platform attribution.

Your business model also impacts your approach to measurement and tools. Depending on your purchase cycle (pure e-commerce, subscription, or longer purchase cycles), you might want to find the right approach to ensure you can measure what matters for you either directly or through proxy metrics.

3 - Your KPIs and success metrics

The third factor, which is related to the first 2, is your success metrics and KPIs. Depending on how you measure your Growth Marketing success (I highly recommend you align it with your activation/onboarding metrics), your approach to measurement will vary. The lower in the funnel, the more challenging it might get to measure your success, and the more sophisticated your Data team would need to be.

4 - Your internal capabilities

Lastly, it all comes down to the capabilities and resources you can allocate to measurement. It's very important to be realistic about the level of effort required and whether it makes sense to follow a specific route. Managing expectations and articulating the tradeoffs is critical to ensure you can keep focusing on the core mission which should be about building a sustainable and predictive Growth engine instead of going into the rabbit hole of attribution and tools if you don't have the resources for it.