You might be wrong if you rely on Google Analytics for reporting the effect of your marketing efforts. In this blog post i take a look at the channel called Direct. – And why it is wrong!
During my work as a marketing consultant, I see a lot of companies that blindly rely on their Google Analytics to tell them how to allocate their marketing budget.
But they don’t know. In fact, almost no one knows that they are most likely looking at the wrong numbers. Read this article to see what I’m talking about.
Most marketing managers are obliged to report once a week, or once a month to their boss; it could be the CMO, who then reports to the CEO or directly to a Board.
Everybody agrees that google is right, and you can trust the data blindly.
So the marketing manager ventures into the google analytics and creates a report of conversions from each marketing channel.
Acquisition > All traffic > Channels >
The CMO is happy to get the report, and reports to the Board who loves to see that the Direct (their brand) is now generating 10,93% of their revenue. (they’re wrong)
This model only contributes value through the last-click model and doesn’t, at all shows the correct data.
The last-click model is a basic model to decide which marketing channel gets “credit” for a sale. Last-click attributes the value for the last channel the user interacted with, before the sale.
Google analytics uses the last-click model as a standard in all the reports, except the attribution models.
Imagine your own journey to a purchase online. Just think about the channels you actively use before buying.
Let’s say you want to buy a new bike helmet.
First you might Google “Bike helmet reviews” and read several review pages on which helmet to get.
After reading the reviews of magazines, you might go to Amazon to see actual user reviews of the helmet as well.
At last you post on your favourite social media, on your smartphone, to ask if anyone you know has a preferred bike helmet, just to hear it from your friends and family.
After deciding which helmet to buy you go price hunting, you use Google Search, Google Shopping, Amazon, eBay etc. and you end up finding two or three online stores to shop in.
But before choosing which shop to place the order in, you go searching for other people’s experiences with shopping at the particular stores.
So now, you have one shop left. To be sure you end up going to your favourite social media again, just to hear if anyone has a great or bad experience with this particular shop.
Meanwhile visiting this social media, on your smartphone, a banner pops up. Amazon has a 50% discount, today, only for you.
You don’t click the ad. But opens your tablet and go directly to Amazon.com
In this example, Amazon would see the sale as a “Direct” sale.
This example is not a fairytale, is actually happens a lot.
Amazon might be a bad example, i imagine that they know how to measure their marketing efforts correctly.
But let’s say that Amazon didn’t know anything else but the last-click model and has concluded that their brand was strong enough to get the sale above.
So now they might cut some of the budget for their remarketing campaign on the Social Media. Which in this case could have resulted in not getting the sale.
The customer journey to completion is so complex, and so unique from user to user, that we have to dig deeper to get a grip on how to allocate the marketing budgets.
The solution to this problem has to come from the top. Because even if every marketing manager learns that the last-click model is outdated and useless when allocating budgets for marketing. It wouldn’t matter if his Boss didn’t share this opinion.
Let’s say we are working with an established xx-year-old brand. It could be McDonalds or Facebook.
They could and would argue, that they have a strong brand, and the Direct channel or their brand-legacy has some value.
Indeed it has, and brands through generations might be able to go on without payed marketing. But nonetheless we have to attribute the credit for a sale to something else, than their brand.
In McDonalds’s case, everybody knows McDonald’s mainly because of two things. Either your parents brought you as a kid or you watch a thousand commercials as a kid with Ronald McDonald.
In McDonald’s case they knew to target kids and imprint it early, so that we’d never forget as an adult.
Facebook is a much younger company, and their user base spread like wildfire. In this case we would attribute the value to Word-of-Mouth and in fact, the marketing cost could be the quality of the product.
In Facebook’s case the knew to invest everything in creating the best product ever. Instead of creating big marketing campaigns, they found out that a great online products helped them grow exponentially
Go to your Google Analytics.
If you have e-commerce tracking set up, you can go to the bottom of the menu and open:
Conversions > Attribution > Model Comparison
And please set the lookback-window to the max number of days available. Right now i it is 90, hopefully they will expand this window.
Then i invite you to try and compare the Last-click (Last interaction) and First interaction model.
In my case, just by looking 90 days back, Direct loses 40% credit for the sales.
If you could go back, to the beginning of your brand and had everything measured through google analytics. Direct would account for 0 sales as the first interaction point.
No one is born knowing your name! Thoughts? Let me and other readers know in the comments.