by Matt Swan
Traditionally affiliates have been categorised by their primary methods of promotion. With boundaries becoming blurred it is important to assess them on their individual merits. It is now difficult to categorise all voucher code affiliates in the same way for instance, as they have different ways of driving traffic to advertiser sites and attract differing audiences.
Advertisers will have a set of key metrics against which they measure the success of their campaign. Typical metrics include but are not exclusive to: percentage of new vs existing customers, average basket values and lifetime value of customers. These metrics will vary from advertiser to advertiser to be in line with their core strategic objectives.
Here we look at an advertiser that sits within the retail sector. The attributes they associate with value include percentage of new vs existing customers, return rate, failed credit check rate, average basket values and most importantly the cost of recruiting a new customer to the business.
New vs existing customers
A common attribute that advertisers are recognising as a key performance indicator is the split of new vs existing customers through the affiliate channel. For this advertiser, internal targets are set on delivering new customers to the business.
While new customer delivery is one indicator of value, other factors also need to be considered. Simply saying the higher the split of new customers the greater the value ignores other metrics that constitute value through the channel.
This advertiser has seen voucher code and cashback sites with a high percentage of new customers. In one instance this is as high as 40% off sales are generated by new customers. This indicates the value these affiliate types are able to provide in terms of new customer delivery, contrary to the myth these affiliate types are more likely to convert existing customers that are looking for a discount.
Returns and Credit Rejects
As the advertiser offers credit accounts, another important metric to consider is how many orders are declined due to a customer failing a credit check. Additionally, retailers will commonly receive a number of returns. A high value for each of these metrics is a negative for the business so need to be monitored closely.
The average returns rate for this advertiser from each of the top 10 affiliates is under 10%. For the top performing affiliates within this category it was as low as 2-3%. Figures for failed credit checks varied widely from affiliate to affiliate. Affiliates with a high rate of failed credit checks could be ranking highly for credit related terms within the search engines. By using these findings, it is possible to tweak the marketing message contained on these sites to minimise the number of failed credit checks. This could be achieved by ensuring that credit terms are expressed clearly across affiliate sites.
It is interesting to note that the affiliates with the lowest rate of credit check failures are cashback sites. This is indicative that customers shopping through these sites have good credit ratings. Additional research has also uncovered that cashback site members tend to have a high disposable income.
Cost of acquiring new customers
The cost of acquiring new customers is the key metric for the advertiser in this case study to determine value. As they only recognise new customer sales internally, their cost of acquiring a new customer also takes into account commission paid out for existing customers. Where an affiliate delivers a number of existing customers spending a lot of money, the CPA of acquiring a new customer will naturally be higher.
Without recognising the role the affiliate channel plays in customer retention, advertisers may see the cost of acquiring new customers as being unusually high where they are paying out a commission for existing customers – however minimal this may be. Additionally, this advertiser awards sales on a 1st click basis internally, however pay out on a last click model which is typical across the industry. In essence, an affiliate can be paid for a new customer sale, while internally another channel is allocated the sale. Again this pushes up the cost of acquiring a new customer – with these sales not being allocated to the affiliate channel while the costs are.
This comes down to a question of where affiliates add value beyond being first click in the sales process. If a customer has not converted from the first click then any additional clicks would have an element of influence over a sale. For example, incentivised sites' business model is to close a sale. Would the transaction have taken place without the aid of a voucher code or cashback site? Looking at the CPA associated with all new customer sales where affiliates have been paid (on a last click basis) could provide a truer reflection of the value they add in terms of new customer acquisition.
Average order value of new vs existing customers
A further indicator of value is how much new and existing customers are spending on average. If existing customers are spending more through affiliates should this be considered when assessing the value of the affiliate channel? It could be expected that existing customers who are already familiar with the brand would be spending more than new customers. This is the case for all affiliates investigated, although for two affiliates the difference was minimal.
Customers are spending more when they are incentivised to do so. It is interesting to note, even though there aren’t discount codes for existing customers for this advertiser, they are still more on average through voucher code sites. Each of the voucher code sites have a higher new customer average order value (AOV) as they are encouraged to spend more in the knowledge they will receive a discount.
Similarly cashback affiliates included in this study have high AOVs for existing customers as they are still able to earn cashback on their purchases – although this is minimal. A high existing customer spend increases the CPA of recruiting a new customer but does increasing existing customer spend add value to the business too? This should also be compared against other online channels to determine whether existing customers are spending more through affiliates than other channels.
The data contained in this document has allowed us to analyse the value each of the top 10 affiliates across a key programme on the network. By looking at the KPI’s that determine the value that each affiliate is adding, promotions can be tweaked to ensure that valuable customers are being delivered while driving the CPA for acquiring these customers down.
The primary objective for the campaign has been to drive new customer acquisition; however this data also investigates the importance of customer retention, and shows how returning customers are spending more through certain affiliates.
It has also challenged the model of how new customers are allocated internally. By allocating commission internally based on the first touch point of a customer, it neglects the influence that other channels may have over a customer. For example, certain affiliate types (voucher code, cashback) prime their business models on closing the sale whereas others (content, comparison) are often used within the research phase of a purchase. Affiliates are able to add value at each stage of the process.
One important metric that has not been examined in this case study is lifetime value of customers. While generating a number of new customers is a key objective, the Holy Grail is also retaining these customers. Affiliates that are not only able to attract new customers, but attract customers who spend regularly should be rewarded accordingly. It is important to benchmark against other channels when assessing affiliate performance. This will allow for more informed decisions based on the value individual affiliates offer.
There are a number of metrics that contribute to delivering valuable customers and it is possible to reward affiliates with individual commission rates based on these metrics.
Matt Swan is client strategist, Affiliate Window & buy.at