How Much is Influence Really Worth?

November 2016

By Anthony Clements, Country Manager

How much is influence really worth? It’s the question everyone is asking right now.

This year I feel like the concept of rewarding publishers for the influence they drive, as well as the conversions they generate for advertisers, has finally taken hold in our industry.

It’s taken far too long really. At Affiliate Window, influence reporting and payments were introduced in 2014. Since then, other UK networks have followed suit. And finally it feels like there is a consensus throughout the industry that influence needs to be properly valued, and rewarded particularly for publisher propositions that deliver value to advertisers but struggle to monetise properly on a last-click CPA.

However, these are still in the early throws of bringing meaningful improvements to a 20-year-old reward mechanism. And that means debate, disagreement and confusion continue to reign. So let’s shed a little more light on the beguiling question: how should we reward influence in the affiliate channel?

Is all influence the same?

The simple answer to this is no. The longer answer involves mixing data that shows how many click based interactions (influence) are delivered by a publisher but not rewarded, with a bit of logical deduction about how influential a publisher’s content is to a user when set against the other channels involved in the chain.

Affiliate Window’s interface publishes a total influence score per-publisher, which is a combination of all click-based influences and transactions. This allows you to work out a percentage influence score for each publisher, which provides an indicative view of an advertiser’s most ‘influential’ publishers.

But the data doesn’t offer the full picture and that is because not all influence carries the same value. Consider this example:

Here is a breakdown of publishers that have overwritten a popular discount code site on the network for a major retail client in a typical month:

The data is a count of the number of times this publisher influenced another, and which publisher type received credit for the final conversion. So for example, of the 1,836 transaction influences made by this discount site, 1,588 of these ended up in transactions for other discount code sites.

Discount code sites overwrite other discount code sites more often than any other publisher type. This isn’t a surprise. In fact, it’s quite normal for publishers to most commonly overwrite other publishers of the same type.

Let’s consider a popular comparison site on one of the network’s biggest telecoms clients.

It’s a very similar story. This comparison site influenced transactions for other comparison sites in 70% of the total influences they were credited for.

Why is this important to understand the value of influence? Because whether it’s influence data, attribution information or intent metrics, there is one important person behind all of it and that’s the customer. Through all this analysis, we are actually trying to apportion value to the importance of online marketing touch points to the consumer. So thinking about the intentions of the consumer is important for deciding the varying value of influence across the different affiliate types.

A consumer using multiple discount code sites during a journey, or more than one comparison site is thinking differently to a consumer that goes via a comparison engine and then onto a cashback site. The consumer that uses two editorial content sites is thinking differently again to the consumer that goes editorial content to cashback site. All of these publisher types have influence value. They all for instance deliver rich and unique content to consumers. Their influence carries different value not because of the type of publisher they are, but because of how a consumer has used them relative to other affiliates.

The value of a publisher’s influence depends on sector, market-position, price-competitiveness, available discounts and crucially how their proposition is being used by consumers. At Affiliate Window, we consider influence payments as a very important method of rewarding some publishers, but influence payments aren’t a box-ticking exercise on the route to innovative affiliate management. They need to be considered in context.

How much should influence cost?

Here are some interesting stats:
Average influence payment across the network: £6.60
Highest single influence payment on the network: £10

Our approach is to think of influence payments as an additional way of rewarding specifically identified publishers or types. Advertisers still pay a CPA if a publisher converts a sale; but rewards them a smaller amount for unrecognised influence if they don’t.

The amount an advertiser pays for influence needs to be proportional to the amount they are willing to pay for conversions.

One advertiser on our network offers £10 per-influence (in-channel only) off an average transactional CPA of £72.

Another in the retail sector offers £1.50 per-influence off an average transactional CPA of £2.10 and an average order value of £126.

A third, high end retailer offers £10 per-influence off an average transactional CPA of £22 and an average order value of £280.

These are indicative examples, but across the network we can observe some trends, such as:

- influence payments are rarely less than 20% of the transactional commission
- very rarely higher than the average transactional commission of the programme
- influence payments are normally between 30% and 70% of the average transactional commission

How is rewarding influence affected by attribution

Plenty of advertisers approach us asking questions about how the affiliate channel should be included in their cross-channel attribution. Particularly when it comes to rewarding publishers differently. Our first answer is always make sure you are using an independent, channel-agnostic attribution partner with flexible technology. Plenty claim independence, but have underlying conflicts of interest.

The second point is that espousing non last-click CPA rewards for publishers does not mean we denounce last-click CPA. In fact, it should remain the principle way of crediting publishers for sales.

We encourage clients to consider attribution modelling as the rule or set of rules that determines how credit for sales is assigned across all online paid media channels. Attribution should provide insight on customer journeys that dictate spend and budget decisions.

Crucially, attribution models should not be the tool by which retrospective changes to the way publishers are paid. This doesn’t happen in any other channel, and shouldn’t happen in affiliates. Attribution models certainly should be used to help identify publishers that could be rewarded for additional influence. This means an advertiser’s attribution technology needs to be able to identify affiliate activity on a per-publisher level. It is also important that all channels are measured equally. So for example, if the action used to measure Display is an impression, then this should be replicated across those publishers that are promoting in a similar way.

Should publishers be rewarded for influencing sales in other channels?

On a general level, yes. Single channel decisions are never going to be as powerful as multi-channel ones. 12% of affiliate channels sales involve more than one publisher, but 62% of sales (based on a major retail client) involve more than one marketing channel.

As an example content publishers, blogs, social-content partners and brand ambassadors should be rewarded for influencing a journey even if brand search is also involved? Influence payments for these publishers is a great way to reward their contribution, especially if a client’s attribution modelling has highlighted that sales are more likely to happen when a customer interacts with these upper-funnel propositions.

Of course, we’ve already considered that not all influence is the same, and one factor that determines the value of influence is the journey type. A content publisher from the affiliate channel influencing a sale converted by brand search for example is going to be valued differently compared with a content publisher that influences a journey where there are three display interactions, generic search, Google Shopping, and finally brand search. Again, that all important context is vital to properly the value of the influencers.

The affiliate channel isn’t far away from getting the influence value of its publishers right. There is now a lot of willing from advertisers, and a desire from publishers to be rewarded for their value beyond just a conversion. The next step is to better understand the differing value of influence across our channel, and to ensure affiliate is properly integrated and valued alongside other marketing channels.

AWIN Strategy Team
Monthly Strategy Newsletter