Having been in the industry for such a long time now and having talked and consulted with many different advertisers in various sectors, I feel there is a still a little mystique to everything within affiliates away from the de facto last click model.
It hasn’t changed in the 20+ years since the start of affiliates: Sales are recorded through the affiliate network, all focused on last click and you’ll see that the program will account for over 15% of the total share of online sales – though I hasten to add this will vary from client to client.
I decided to investigate a different perspective; one that doesn’t allow last click to be the main KPI (Key Performance Indicator), as shown in the bullet points below. This is a client of ours, who tracks on a popular affiliate network and uses Google Analytics to record all online sales – affiliate marketing to them according to the network last click figure accounts for just above 18% of all sales. However, by looking at proprietary technology at our disposal, which looks at every sale and it’s full customer journey from start to finish, showing all touch-points, we are able to ascertain.
- Affiliate last click sales actually accounted for 10.8% of total sales revenue in June 2017. The reasons why this is lower is due to genres like display not being the ‘true’ last click, as well as other partners who may have been de-duped in the process. More interestingly;
- 8.8% of all sales occurred through affiliates on a first-click basis. This actually demonstrates a strong level of value from the channel at the beginning of the customer journey. This is one side of the channel that is seldom seen, and actually, we see a lot of strong content sites, bloggers, and influencers in this cohort that does not get rewarded.
- 2.6% of all revenue generated involved affiliates at some stage in the customer journey where they were neither the instigator nor the closer (First click/last click) and therefore were not paid or acknowledged when purely looking at the last click model.
- The total percentage of sales where affiliate initiated and closed the sale was 5.5% – this is one touch, one conversion. In and out. Not bad going. This accounts for just over half of the last click sales (see first bullet point)
- The most lucrative customer journey was a two –step journey. This didn’t include affiliates but was ‘SEO > Direct’, which had an average order value of 151% higher than the average. Breaking down each sale and seeing the path this way allows us to recommend how to get more of the same sale (E.g. invest more in SEO keywords that are working, optimise website UX for these products, optimise product feed, etc…)
- Total latency of an affiliate sale was 10 days on average; whereas direct, it was a more considered purchase with double the latency, 23 days. With incentive sites being prevalent on this program, perhaps that indicates that a small discount ‘seals the deal’ early.
I’m hoping you will agree that these metrics provide a much more rounded version of events of how the affiliate channel assists, rather than acquires. And to me, that’s the hidden value that marketers need transparency on in order to construct truer stories on how the budget is best apportioned for not just the affiliate channel, but across all channels –a multi-channel strategy if you will.
Just to take this up a level it got me thinking about what specifically we should be recording moving forwards as an industry. Below are my thoughts:
The Last Click Model – We’re an industry focussed on the last click model. Now, I’m not saying whether it’s the right or the wrong way to accurately measure the channel, but we shouldn’t disguise or ignore other stages of the total customer journey, as shown above. I would like to see how affiliates help with the start of the journey, particularly if it’s a high-ticket item. For those identified as key first click initiators, we can start investing in these partnerships on a different deal away from the last click – perhaps a small bonus or tenancy, or perhaps something a little more. To use a loose football analogy, instead of affiliates predominantly being goal scorers (for example, incentive sites closing the sale), we need to become more focused on the run made by the right wing-back that assists other channels or may actually score a goal themselves.
More Meaningful Metrics – Taking the above one step further, there are other metrics that we feel should be recorded within the channel, such as:
- What channels are over/under valued (The incrementality/“would I have got that sale anyway?” debate);
- ‘If click’ (what channels interacted in the path to sale), and;
- Profit margin – broken down by most profitable items sold on site. How as a channel are we able to influence more of the same?
Every advertiser will have different metrics to report back on, but a higher amount of consideration should be given to those that will ultimately reduce cost and help with constructing a multi-channel approach, rather than simply treating affiliates in isolation.
The Perfect Customer – The affiliate channel has always been strong in delivering new customers, and these parameters can be passed through the majority of affiliate networks. We’ve been focusing on this as well as improving the value and lifetime of the customer through data, targeting elapsed customers as well as testing promotions up and down the funnel depending on what stage of the purchase the customer is at: attention, research, or purchase. This was covered earlier on this year in how, “Seven is the magic number of customer touch-points for optimal average order values”, so do take a look if we wish to learn more about gearing up activity for high AOVs (Average Order Value).
Affiliates as “Free Advertising/Branding” – Affiliate is one of the only channels that don’t get rewarded for anything but the conversion: PPC – you’ll pay for each click; Display/programmatic, you’ll be paying on a CPM (Cost per Mille) basis; Email on a CPM also, etc…, which for advertisers is great, but is a hidden value that is often forgotten and is worth repeating.
As per the example noted in the real-life example, 8.8% of sales in the example were initiated by affiliates but were not necessarily rewarded for the sale. The industry often writes about the strong ROI of the channel (£13 return for each £1 spent), which makes me think that there is a spare budget from advertisers to fuel the top of the funnel with tenancies or branding activity. This can only be done if recording the first click or clicks on the run up to the last.
Remember: You can only manage what you can measure.
Latency – Recording the latency of all sales are becoming more important for marketing managers as a quicker sale could mean a cheaper total cost of sale. For example, the fewer PPC ads engaged with, the less paid social ads clicked on, which will ultimately mean less money spent on acquiring that customer altogether. Affiliates are very good closers and can be highly impactful on how quickly it can close based on the value of a discount or the call to action of the product in question. From here, and from knowing latency, campaigns can be A/B tested with different partners with various promotional values to see what works, what doesn’t, and improve the value of the channel moving forwards.