How to measure the ROI of your affiliate marketing

17th Aug 2015

Affiliate marketing is unique in being the only manifestation of the discipline in which everyone measures at least some kind of data.

They have to because, as Helen Southgate, UK managing director of affiliate network affilinet points out: “Payment is made on a per sales basis so it’s requisite that brands measure activity or they wouldn’t know what to pay.”

But despite this situation, an econsultancy report indicates that a worrying three quarters of organisations currently fail to consistently translate any of their marketing data, whether affiliate-driven or otherwise, into actionable insights.

Susan Hallam, managing director of digital marketing agency Hallam Internet explains the significance: “Affiliate marketing generates vast quantities of data, but the big question is how people then act on that data. So if they establish whether this ad is more effective than that one or if that partner is doing better than the other, what are they actually doing about it?”

The idea is to use such data in the most effective way possible in order to optimise performance. Here we explore how best to go about doing just that:

1. How will I lose out if I don’t measure my affiliate marketing activity?

Big High Street brands and companies that generate a hefty percentage of their revenues from affiliate marketing tend to have an in-house team to manage such activities, while others often rely on digital marketing agencies or affiliate networks. But whatever their chosen approach, they all still need to be on top of their data in order to make informed decisions.

If they don’t, according to Jacqueline Cox, senior affiliate marketing manager at marketing and technology agency DigitasLBI, the danger is they simply will not get the most out of their investment and will miss potential opportunities.

“It won’t be total death, but if your affiliates aren’t really engaged, they may lose interest and go to rival brands who are giving them loads of cool stuff to put on their site. This means that you’re losing opportunities to drive more sales,” she explains. “Also any information is only as up-to-date as the most recent information that you’ve sent them so if an offer runs out, you may end up with disappointed customers.”

Another consideration is not getting value for money. “With affiliate marketing, you generally only pay if someone buys something, so it only costs you if your affiliates make a sale,” says Hallam. “But all the while you’re still doing artwork, making offers and the like. So if you’re not measuring performance, you’re genuinely wasting money.”

A further consideration here is that, while the majority of affiliates will not generate much in revenue terms, a small percentage of “super-affiliates” will. “But if you don’t measure, how will you know who your super-affiliates are and how they’re performing so that you can harness their energy, trust and goodwill,” she adds.

2. What should I measure?

Here are some examples of common key performance indicators (KPIs):

  • Revenues or sales – how much money you made in total.
  • Conversion rates – what percentage of people an affiliate encouraged to visit your website who subsequently purchased goods or services.
  • Cost per sale – how much you had to spend to make each sale happen.
  • Return on advertising spend – how much money you spent on a campaign and what your subsequent margin was.
  • Average order value – how much an average order is worth.
  • Affiliate behaviour – understanding how many affiliates you have, how active they are and how much revenue each one generates.
  • Average earnings per click – this figure can be calculated by dividing the total commission an affiliate earned by the number of clicks they generated to help you see whether they are providing value for money.

Another newer KPI that is currently only used by about a quarter of advertisers is customer lifetime value – the idea here is to track not just sales acquisition-related data from the affiliate side, but to combine it with post-sales information from the brand side, generally over a 12-month period. Such information includes churn rates, how much each person spent on their site and whether they were a new or returning customer and taken together shows just how valuable, or not, a given customer is to the business.

3. What tools and services are there to help me measure effectively?

  • Affiliate networks – the three biggest in the UK and US are Affiliate Window, CJ Affiliate and Tradedoubler. The networks recruit and pay affiliates on behalf of brands, while also providing brands with access to dashboard-based interfaces of varying degrees of sophistication so that affiliate performance can be tracked. Results should be monitored daily, or at least weekly, to ensure advertising programmes are meeting expectations. If advertisers’ campaigns make significant amounts of money, the networks also supply them with a dedicated resource manager, but if not, support will be more limited to activities such as trouble-shooting. Brands will be charged commission and an override fee, which pays for the administration of the affiliate network
  • Digital marketing agencies – these provide a full-service offering, which includes a daily or weekly performance report.
  • Tools to help content generation - Buzzsumo, for example, enables you to see who are the most influential people online and what they are talking about. It also generates potential ideas for bloggers or other content affiliates. Google Keyword Planner enables you to look at what search terms people are using most on Google such as where to buy BBQs if it is a sunny day in order to help you plan campaigns around them.
  • Click-fraud and policy enforcement tools – products such as The Search Monitor and BrandVerity enable you to monitor brand and trademark use and prevent trademark poaching. The aim of using these tools is to reassure affiliates that they are being fairly compensated for the sales they generate.

4. What key challenges am I likely to face?

One of the biggest challenges with any form of digital marketing is the sheer amount of data generated by such activity. This means that organisations often do not know where to start in terms of analysing it - particularly if they have no coherent affiliate marketing strategy in place - or do not necessarily have the skills in-house to do so.

The flipside of this is not making the most of the data you have. Affilinet’s Southgate explains: “In digital marketing, we’re very guilty of the fact that because we can track so much, we want to have everything rather than thinking ‘what have we got and how can we use it in the best way?’”

It is also worth bearing in mind that, when launching an affiliate programme, it can take a good year to see meaningful trends in your data as statistics have a habit of going up and down so you need to bear with it.

Other thorny issues though, according to Chris Kramer, partner and head of operations at digital marketing agency House of Kaizen include marrying affiliate networks’ marketing data with brands’ own client sales and post-sales data, which is often scattered across multiple systems, in order to measure customer lifetime value - a process he describes as “difficult but not impossible”.

Just as challenging is dealing with attribution and the role that affiliates play in converting potential customers into buyers. “Depending on how many channels you’re measuring and what type of order volume you’re dealing with, the complexity of the data makes accurate analysis difficult,” Kramer says.

5. What are the critical success factors for getting it right?

Forward planning is vital if brands want to effectively measure and act on the performance of their affiliates. So this means comparing any data against a 12-month strategic plan in order to ensure that activity is on track. As Hallam points out: “You’ve got to take your affiliate marketing as seriously as you would business planning.”

Another consideration is to “test, learn, test”. “You never know which advert will be best or what offer will hit the spot so you need to be testing all the time and learning from it,” she says. “There will be campaigns that you know work and they’ll be your bread and butter, but for the rest, you have to be agile.”

But it is also important to ensure that you are able to maximise your sales and get value for money out of your budget. To do so involves setting benchmarks as to how much you are willing to pay for activities such as cost-per-action and continually running scenarios in a tool such as Microsoft’s Excel spreadsheet.

“You need to constantly weigh up your budget versus what you can get out of it – so how much per sale are you willing to pay and don’t go above that ceiling? The question you should constantly be asking yourself here is, ‘is this a good investment?’” concludes digitaslbi’s Cox.

Replies (1)

Please login or register to join the discussion.

Dr Andrew Lancaster, Director of
By Andrew Lancaster
11th Jan 2018 20:09

To use data in a productive way, it often helps to start with a question (e.g. how much of our revenue can be attributed to the affiliate program?). Then you can start to use the data intelligently. You may find you have critical data gaps.

Thanks (0)