Marketing attribution: two words that can send shivers down the spine of even the most seasoned marketers.
Marketing attribution is like a detective game. It’s about figuring out which marketing moves really helped your startup grow. You have to get many things right to make it work.
The people who care most about this are the marketing team and the bosses. They want to know what’s working in their marketing plans so they can do more of it.
The journey of building the attribution models was far from straightforward. The knowledge we gained has been instrumental in reshaping our approach to marketing strategy, allowing us to make more informed, data-driven decisions.
Table of Contents
The Early Chaos: Fundraising and the Wild World of Marketing
Picture this: we had just raised our Series A money, and it was exciting! We were talking to big influencers, planning TV ads, and even thinking about sponsoring an F1 team. We were ready to spend a lot on marketing. It felt like the wild west!
Campaigns were happening everywhere. We teamed up with famous YouTubers, but we didn’t really know how well these efforts worked. The marketing team kept calling about coupon codes on weekends as we built a big customer funnel.
One advantage we had amidst this chaos was that the entire data engineering and analytics engineering layer was set up and running by now.
We had chosen the modern data stack and were using tools like segment, DBT, appsflyer for a lot of data tracking, data transformation aspects, which made life easier and data systems more robust in the medium to long term as I’ll describe below
The Art of Inquiry: Identifying Key Questions to Drive Marketing Attribution Clarity
To make sense of this chaos, we had to gather some key ingredients by asking the right questions. I’ve emphasized on this most of my data career and some of my previous case studies also prioritize this. Another popular approach is the Crisp – DM framework, that can be used well here too, which follows similar steps elaborately.
We started by asking our leadership, marketing team, content team and the technical team the following questions to understand which ones were the most important conversion triggers and how best we could reverse engineer them via our existing data architecture and customer journey. These discussions were crucial not only in gathering key ingredients, but also aligning all stakeholders together to a unified attribution model and strategy as we built it.
Marketing and Content Team:
- What counts as a conversion?
- How many times does a user need to see our ads before they install the app, sign up, deposit money, or come back after leaving?
- Which interactions are the most important?
- How can we tell if a campaign meant for new users also re-engages old users?
- Where are we losing users in our signup flow?
Technical Team:
- When do user actions trigger events like install or signup in our system?
- How can we smoothly connect our data with other software, like Appsflyer?
- Can the recorded install time be wrong?
- How strict should our attribution models be?
Our Plan: Short-Term Wins + Long-Term Success
We needed to balance immediate needs with long-term goals. Here’s how we did it:
Short term strategy – First-Touch and Last-Touch Attribution:
- We started by implementing both first-touch and last-touch attribution models. Think of it like this: first-touch is like remembering who introduced you to your favorite band, while last-touch is recalling the final push that made you buy concert tickets.
- First-touch attribution gave credit to the initial interaction, helping us understand which channels effectively initiated customer journeys.
- Last-touch attribution focused on the final interaction just before conversion, providing insights into the channels that sealed the deal.
Long term strategy – Multi-Touch Attribution with Points System:
- Our ultimate goal was to develop a multi-touch attribution model that fairly distributed credit across all customer interactions.
- We designed a points-based system to weight different touchpoints, offering a more comprehensive view of the customer journey and the relative influence of each channel.
It was time to build. The tools, tech stack keep evolving so they aren’t the most important thing, the approach and the solution design is crucial. But, if you have the right tools and you want to buy vs build for a rapidly growing startup, then our tech stack that we implemented would be quite useful for your use case.
Tech stack for Attribution and Data Collection:
- Appsflyer: Used for people-based attribution, primarily leveraging last-touch models.
- Segment: Captured frontend events and clickstream data.
- Snowflake and DBT: Formed the backbone of our modern data stack for data warehousing and transformation.
- Backend Attribution: Implemented via joining codes distributed to marketers.
A crucial part of our strategy was to finalize the conversion points and rigorously test them across various systems:
- Frontend Events (Segment):
- Mapped anonymous IDs to user IDs during the conversion process.
- Ensured that events such as “user_signedup” were accurately captured and attributed.
- Backend Events:
- Verified the occurrence of sign-up events in our backend systems.
- Appsflyer Integration:
- Ensured Appsflyer accurately identified the correct source for conversions.
Aligning these data points required significant coordination and testing across technical teams. This alignment ensured consistency in our datasets and attribution results across internal and external sources.
Practical Steps and Testing
- Define Conversion Points:
- Identified all possible verification points for a visitor becoming a customer, both in internal systems and external software.
- Test and Iterate:
- Conducted rigorous testing and back-and-forth coordination between teams to ensure data consistency and accuracy.
Our journey taught us that building a great attribution model isn’t just about fancy tech or complex formulas. It’s about understanding your business, being patient, and not being afraid to start small.
So, if you’re about to embark on your own attribution adventure, remember: start simple, dream big, and don’t forget to celebrate the small wins along the way. Who knows? You might just revolutionize how your company views marketing!
Ready to Take Your Marketing to the Next Level? Has our journey inspired you to dive deeper into the world of marketing attribution?
I’d love to share more insights and help you navigate your own attribution adventure. Whether you’re just starting out or looking to refine your existing strategies, I’m here to help.
Book a free consultation with me, and let’s explore how we can tailor these strategies to your unique business needs. Together, we’ll craft a roadmap that turns your marketing data into a goldmine of actionable insights.
Leave a Reply