Marketing Attribution is a mess. There, I said it! 🤷♂️
First, what the hell is Marketing Attribution?
Marketing Attribution is what we call understanding what sources and touchpoints have generated a specific outcome—for instance, a lead into the pipeline or a sale in your e-commerce store.
You would think that within the Digital Marketing realm since we’re able to measure pretty much everything, it would be a walk in the park.
Well, not so much.
Because data has flaws originated for three main reasons: Reach, Correlation and Expertise.
1️⃣ Reach — We can measure what we can measure. 🤷♂️
Chris Walker and the talented people at Refine Labs have been championing this claim that you can’t correctly attribute revenue to particular sources because we can’t measure things like word-of-mouth, podcast influence and many more channels or tactics.
They’re absolutely right. You can only present what you have data to show in a pretty dashboard. Since you can’t find all possible ways of tracking how people share information and talk about you and your brand, you can’t attribute actions to the outcomes.
2️⃣ Correlation — When in doubt, the following data gets the credit.
When you can’t reach all of the touchpoints that originated the outcome, the last original data you collected get’s the praise.
Imagine what happens when someone searches your brand name because they’ve liked your webinar and want to know more about you.
Yes, Google becomes the source.
I see this happening in many email marketing and marketing automation audits. Sometimes, email marketing gets the credit for a sale when they’re just collecting what your branding and advertising ads already set the desire. We’re just there at the right moment to collect. 😉
3️⃣ Expertise — “Technology is only as good as the fools who use it.”
I think Aaron Heckhart drops a line like this in one of his movies. 😅
Basically, even if we have all the data sources and tons of reports, it only makes sense to people with enough know-how and experience to extract insights from that.
I should know. People pay me big money to tell them what the data shows me and how they can improve their email marketing and automation efforts.
That’s why Chris and his team ask their fresh leads where they have known them in the first place. Eventually, they can even ask them in a conversation what were the things that made them trust enough to reach out.
It’s like a hack to have at least some sort of insights to make decisions and decide if it’s worth it to keep with the podcast or other tactics that you can’t measure directly.
Yes, the first channel I will talk about is Email Marketing Attribution. 😅 Mind-blowing, right? 🤯
Email Marketing Attribution
So, let’s start running: Email Marketing is one of the best-performing channels in your Digital Marketing stack. Period!
That being said, email marketing benefits greatly from all of your other channels and especially from your brand and rapport/trustworthiness built over time.
Here are a few things that may impact how you measure email marketing against other digital marketing tactics.
1️⃣ Different Platforms measure things differently.
Let’s say you have your email marketing platform connected to your e-commerce store. If someone goes from your email into your e-commerce and buys something within a particular time window, it will attribute it to email.
On the other hand, if your Store Analytics collects other trigger actions from other sources like retargeting ads, and the last click comes from these ads, it will attribute it to retargeting.
In reality, the attribution of that sale should be to branding, performance advertising, retargeting ads, influencers showing off your product, your offer pop-up, your recovery cart tactics, and so much more.
2️⃣ Reports lie.
There, I said it. 🤷♂️
Reports show what your platforms can measure and your team can extract from it. It doesn’t tell the whole story.
For a layman, these numbers can represent absolute truths.
The email has generated 80% of your revenue. Long live email.
The problem is that decisions are made based on these reports.
If you trust these numbers, what would you do?
You’ll eliminate those other tactics that are not getting revenue and double down on what it does.
3️⃣ Last-Click vs Multi-touch Attribution.
Due to tracking constraints and privacy concerns, attribution is a complete mess.
The simplest thing to do is to attribute a sale or a lead to the last click which generated that conversion.
However, as you may already see above, you’re discarding a lot of other touchpoints within your buyer journey.
Tracking all of the other touchpoints in your customer journey is challenging because most happen outside your web properties.
➜ It’s the engagement and impact in your social media posts;
➜ It’s the targeted advertising within your ideal audience;
➜ It’s the influencer that shows how your product looks good on them.
That’s when expertise is critical. To have people inside your team or consultants who can look at your data and your investments and determine how these actions impact your revenue.
Email is an excellent finaliser of your digital marketing. Consider it the wide receiver or the striker of your entire football team. But without the team to give them the ball, they can’t score.
It’s not magic. You can measure things that influence your bottom line that are not in your usual flow. 🔮 👇
How to Measure Things Without Measure?
If you’ve been following this week’s In-Depth How-To Series on Marketing Attribution, you may already have an idea of how much can go wrong.
Decisions are made based on what’s on record.
The problem relies on what you’re not measuring, which affects how you make decisions.
“Ok, Rui! Enough of this introduction. Give me the “juice”!” —Your little brain cells scream out loud.
Wait, I know, it’s obvious. Yet, it needs to be here.
You can ask your new clients, leads, or new purchases where they have heard about you or what triggered them to opt for you instead of the competition.
The way you can implement this is by simply adding a new form field asking where they heard about your brand. Inserting in the script of your onboarding process accounts is also a must.
Wait, I’m going to explain in plain English. 😅
See those highs and downs in your new business record charts? If you can’t see them, you’re too close. Zoom out!
Give it a broader time span—for instance, six months or one year.
See it now?
Those bumps may be affected by something you’ve launched or removed a few weeks before. A podcast, webinars, branding and awareness campaigns come to mind.
Now compare it with the previous season. Please give it some context (Ukrainian invasion, COVID, etc.).
Dial in what you’ve done within that time and see the correlation with the results you’ve got.
There’s tons of data passing through your digital and web properties. Most of that is scattered; even if it’s collected, it’s not “worked out” to provide insights.
A few examples:
➜ Your forms collected hidden fields that may provide more insights. For instance, the URL source of that lead.
➜ Your client has a geo-IP from the US but accesses from home in Paris. He might have a VPN, and that’s why your emails were not getting enough traction due to the scheduled sending time.
➜ You tagged a few new clients as direct, but your referral shared they’ve mentioned your brand in an event to a crowd of your ideal audience. You might have a few ones that have originated from this approach.
Here’s when you need to put on your Sherlock Holmes hat and scan the traces for clues.
Why shouldn’t you believe in Martech Reports as written in stone?
Yes, every Martech vendor will tell you how their solution is the greatest, and you can’t live without it.
They’ll prey on your hunger for ways to accomplish your growth objectives.
➜ Do you want to do that at scale? You can’t do it without this (expensive) software;
➜ Do you want insight into where your best customers are coming from? You need this over-the-top solution;
➜ Do you want to grow 10X with your digital marketing initiatives? Enrol on this one-year contract for 5-6 figures;
🙌 The thing is, technology is only as good as the people behind it, and even then, it has flaws you can’t solve.
Let me explore the ways martech lies:
7️⃣ It can only give you the findings of what they can measure;
Have you seen the previous posts about this week’s series on attribution? Yes, some things are out of their reach, so their insights are biased toward what they can measure.
8️⃣ Even when they can measure, they are biased;
Have you noticed how your recovery cart automation platform registers much more sales attributed to your flows than the ones your Analytics show?
Yes, the tracking is there, but they account for more sales from their source. Well, they need to justify their fee, right?
The same goes for retargeting platforms and so many others. The discrepancy between what the Analytics counts as visits against what your traffic sources claim is one the best known.
Yes, I know the definition of a new visit differs, depending on your Analytics, but still. You’ve been charged by 2X more clicks without any result to show. 🤷♂️
9️⃣ Integrations have flaws in it, and they know it;
One of the best things about digital platforms is that we can —in theory— integrate them and let them communicate to run more complex tactics.
Well, not so much when some of the integrations are not transparent, causing faulty results. When we forensically notice that and research the causes, in very tiny and shy helpdesk documentation, it’s referred that these integrations have limitations.
If it’s because of one part or another, they don’t say, but the integration can provide inaccurate results, and it’s like the price of doing business. 🤷♂️
Look, as in everything in life, you need to get the results with a grain of salt and not trust what the data tells you until you have done your due diligence and forensic work on those numbers.
Guess what? It’s why people hire my teams and me. 😅
How Marketing Attribution Influences How You Make Decisions.
🤔 Today’s marketers are, in some ways, a “jack of all trades”.
➜ Understand the math behind the investment, results and impact on the bottom line;
➜ Understand how to use the different channels to do that in a cohesive approach;
➜ Understand how to make your brand’s messaging relevant;
➜ Understand the psychology of your ideal audience;
➜ Understand business;
➜ Understand people;
➜ Understand coding;
And, besides all of this knowledge, they’re dismissed by the management and other departments in terms of “having a voice”.
While walking on this thin ice, every decision on where to invest your team’s time and budget must be thought of carefully to avoid being thrown out to the lions if the numbers don’t match.
That’s where marketing attribution seems like a life vest to every CMO. They can always point out what the numbers say if something goes sour.
Data drive us.
Data is like a warm blanket that protects us from the cold weather.
Data seems safe —until it isn’t.
If all we need is data, we could let AI and quantum computing do our work and might as well change jobs.
Yes, data should give us pointers.
Identify trends and alert us from things unseen by the human eye.
But data is only as good as the context, collection accuracy and how you work the information.
We can have a statistical report and still come up with ten completely different insights and outcomes.
➜ Here’s my take on Marketing Attribution.
It’s not in a particular order.
1️⃣ Marketing Attribution is an excellent indicator but should not be definitive in a marketing decision;
2️⃣ Surround yourself with people who can see through the data and come up with real outtakes;
3️⃣ Make sure you deeply understand how every platform measures the same categories/topics;
4️⃣ Insert as many touchpoints and inputs you have into the data analyser for better accuracy;
5️⃣ Deeply understanding of the indirect correlation between your different channels;
6️⃣ Have a hybrid model of tracked data, qualitative analysis and human feedback;
7️⃣ Question yourself, if you hadn’t this data, what would you do instead?
8️⃣ Ask an external consultant for an unbiased opinion;
9️⃣ Take all the results with a whole bag of salt;
🔟 Zoom out and try to see it all;
Bottom line: Your expertise will define how much of this data is relevant to your strategy.
➜ Hey, don’t want to miss the other posts I’m preparing? Sign-up into my newsletter then. It’s quite easy!
Leave a Reply