What is the biggest attribution mistake that most brands make? Agonizing over “what channel gets the credit” for a sale.
This content was originally published in the No Best Practices newsletter on 07.30.2023.
The most common mistake I see DTC brands make when it comes to marketing attribution: trying to shoehorn their attribution model into a Google Analytics-style “what channel gets the credit?” framework.
Why is this a problem? Because consumer demand is an ecosystem, and different channels play different roles in influencing that ecosystem. In fact, many of the causative factors are random and entirely outside your control.
When you’re zoomed in on “what channel gets the credit?”, you tend to measure the wrong things. What you should be asking yourself is “I think that if I do A, the result will be B. Is that accurate?”
Here is a simple example:
A brand has two main marketing channels: Meta ads and Google ads. Their attribution tools says that a specific Google campaign drove $10k in “attributed” revenue, while GA last click says it only drove $5k in revenue.
Seeing this, the media buyer decides to scale the campaign. Three weeks later, things are looking great in both the attribution tool and in GA–attributed revenue is up in both places.
But…new customer revenue is flat, and overall sales growth has slowed down. That probably wasn’t the result that the media buyer, or the management team, was looking for.
What Is My Goal?
If you are a brand doing under $50M in sales, your goal should be customer acquisition. If you have a dollar to spend, it is almost always going to generate a greater return influencing a new customer conversion vs a repeat buyer conversion.
The reason? You have a limited ability to drive incremental spend from repeat buyers.
When you’re looking at your media investment, you are trying to accomplish two things:
- Acquire more new customers at whatever cost the business deems acceptable
- Reduce non-incremental spend
As a result, you should be asking two questions:
- If I increase spend here, will new customer sales increase (while remaining within my KPI)?
- How much of these sales would have happened in absence of my media investment, and how can I take action on that info?
For small brands–especially those using Meta ads as their primary new customer acquisition vehicle–the spreadsheet is the best way to answer these two questions.
How to do it:
- Make one major change at a time in terms of media investment.
- Determine if business-level new customer sales increase.
- Check to make sure your CAC:AOV ratio and MER are still acceptable. Rinse and repeat.
You need to start with a solid hypothesis for the role your new channel or tactic is going to play in the ecosystem. This will help you set a reasonable timeframe for evaluating the questions listed above.
Channels and tactics that access in-market demand (Meta conversion campaigns, Adwords non-branded search) should start to pay off almost immediately. If you aren’t hitting your KPIs (or close) within a week, something needs to be recalibrated within the campaign.
Other channels and tactics build awareness with cold audiences and drive them into your cookie pool. The payoff period on these tactics is going to be longer, because you have to wait for members of those audiences to shift “in market”.
But My Brand Is Complicated
It’s true: for larger brands, things become a bit more complicated. How do you isolate the impact of an increase in Meta spend when you have a product going viral on TikTok, five new stores opening and celebrities wearing your brand on the red carpet?
The goals can also be a bit different. Billion dollar brands can realize more upside from marketing to existing customers, and they might opt to do just that in order to seize or defend market share from the competition.
But the framework is still the same: I took action A. I expected it to cause result B. Did that happen?
Large brands with complex, multi-channel media strategies will often use MMM models and related experiments to accomplish this. This is essentially a statistical model that allows you to understand the causative effect of different aspects of your marketing mix.
Analysis Paralysis: The Biggest Attribution Mistake
In one of my first eCommerce jobs the leadership team was obsessed with understanding the “customer journey”. They wanted to map every click and every view from every customer (impossible) so that we could really understand what was “causing” the sale.
As you can imagine, this impossible quest resulted in very little action or meaningful change. We were supposed to hit a blended last-click ROAS target of 3x with our media spend. This made management wary of investing in higher funnel channels. Eventually our leaky bucket ran dry and sales flatlined.
You cannot eliminate all uncertainty from your marketing mix. Media is never solely responsible for “causing” the sale. Stress in your customers’ lives “causes” the sale. Childhood trauma “causes” the sale. A conversation with a friend “causes” the sale.
Your media helps remind potential customers that you exist. It teaches them more about you. And sometimes it puts the right product and the right messaging in front of them when they’re already highly motivated.
When marketers get wrapped up in debates over which channel “caused” the sale, they often lose sight of the forrest by zooming in on the trees.
Although it is possible to make your marketing mix more efficient, it’s impossible to make it perfectly efficient. Don’t let the perfect become the enemy of the good.