Channel attribution: It doesn't have to be a marketing mysteryby
Channel attribution is seen as very complex and it is, but only if you make it so. When broken down it’s really quite simple. Track and add a monetary value to the sources that bring you revenue. In other words, look at the profit of each channel.
So, are the companies that specialise in attribution making the process deliberately complex and “scary “so you’ll pay them to do it for you? Yes, I think they are.
There are many different models to attribute credit to a source and listed below are just some of them.
1. Last interaction/last click attribution model
2. Last non-direct click attribution model
3. Last AdWords click attribution model
4. First interaction/first click attribution model
5. Linear attribution model
6. Time decay attribution model
7. Position based attribution model
Having all these models to choose from can be daunting. However, once you’ve got to grips with each one, it’s really just a case of applying common sense when choosing which is right for your business.
Getting set up is a bit of a pain, it’s true, and if you don’t have a team of developers it can be a little confusing and time consuming. However, it is not impossible and it’s definitely worth the hassle.
So how does it fit within an affiliate marketing campaign?
Some merchants in the affiliate marketing arena will already be using attribution. Those that aren’t should be. Put simply, if you are not sure what works and what doesn’t for your business then how do you know where to spend your marketing budget for next year? Guessing at this stuff is not an option. It is a recipe for disaster and with many things being digital these days there is really no excuse for guesswork.
The fact is, channel attribution not only allows you to see where sales are coming from, but also identifies which areas are failing and which have potential.This data diversifies your marketing, reducing the risks for many factors such as Google algorithm changes, penalties, PPC price increases and so on.
There are a number of different attribution tools on the market, as well as specialist companies such as TagMan and Intelligent Reach if you’ve got the budget, but don’t overlook Google Analytics (GA). It’s free and has some fantastic tools that will work for small to medium companies.
GA does have its limitations. The multi-channel conversion visualizer on the overview section and the path length tools are great as visuals, BUT, you can’t take any action from this information. Even if you know what path people are taking to make the conversion, be it SEO, PPC or referral, there is nowhere to go from there. After all, you can’t force a person to take a different route.
However, there is a 3rd tool within GA called the Model Comparison tool. This allows you to attribute credit to all your marketing channels. If you use this, a word to the wise. The results will depend on the attribution model you use and as I mentioned above, there are many.
Briefly, the models allocate credit in different ways to the source. So let’s look at each of them in a little more detail:
Using this path as an example - referral >ppc > social media > organic> direct >
1. Last interaction/last click attribution
Would give all credit to direct leads, completely ignoring the upsell and brand awareness created by the other channels.
2. Last non-direct click attribution
Which GA uses as default, would give all the value to organic leads as this was the last “campaign”. I think this is the old school approach. Why would you devalue the direct traffic? Search is becoming more and more about the brands and we should recognise this.
3. Last AdWords click attribution
Does what it says on the tin, gives all the credit to PPC – This doesn’t make any sense at all and, I think, should be ignored.
4. First interaction/first click attribution
This is the total opposite to the 1st model. Thisgives all the credit to the first interaction, which in this case is the referral – but why? It didn’t convert, it was a factor in the sale but the direct traffic made the client purchase.
5. Linear attribution
This gives equal credit to all factors, this is OK but, as old school reports would say, must try harder.
6. Time decay attribution
This is the best option. This spreads the credit, in our case referral 15%> ppc 17%> social media 19% > organic 23% and direct 26%. Put simply, this model just uses common sense. The further back the touch point the less credit, because if they were fantastic, they should have converted right?
7. Position based attribution
A great model however, unless you are a technical expert, it’s extremely complex for anyone starting out.
As a business new to attribution and with limited expertise and budget I’d suggest that model 6, The Time Decay model, is a great place to start.
An important point to remember is that the conversion doesn’t have to be a sale. It is whatever is important to your business. I’ve heard many companies say “we can’t track conversion because we are not an ecommerce store”. That’s just not true. You can track enquires, newsletter sign ups, call back requests, anything. These all have value.
So what effect will channel attribution have on affiliate marketing? There are two answers to this.
Large merchants (and little ones for that matter) should be using this, but very few are. As a merchant you would be able to see the path a customer took and at what point an affiliate was involved. Is that interesting? Some would say yes, very, tell me more and others would say who cares if we made a sale!
I can understand both points of view BUT if the path is >PPC>Direct>Affiliate>PPC> Direct> the affiliate site didn’t introduce your brand and didn’t convert on its own. That’s very true but it certainly helped. Why wouldn’t you like to know the value of each? I know I would, after all I am paying them for the sale. Additionally, you would also be able to see if a particular path was converting the most, perhaps PPC introduces and an affiliate closes? Or vice versa.
Some would argue that this maybe too transparent particularly if a merchant uses the wrong model and gives all credit to the first or last touch, meaning the affiliate has no value in that case. But if the merchantuses the right model the affiliate is adding value. As an affiliate, you should be tracking your own visitors too.
In summary, attribution a great thing and everyone should spend time looking in to it and assessing how it can benefit them. Don’t be lead by the big providers or automatically hand it over to them because it seems too complicated. It isn’t. Do a bit of research yourself and use the information wisely. Not only will you be very pleased with the results, you could also save your business a great deal of money.
Kirstie Eager is head of client services at MoreNiche.