One question we receive is why does Extole need all purchase information. Why can’t I just send the referral orders and not send the rest? Our objective with the Extole referral program is to maximize the number of advocates and their level of activity. We seek high participation and high satisfaction. Conditional firing, or the selective passing of conversion events to Extole, has been demonstrated at a significant number of clients to lead to under-recognition of referrals which leads to low satisfaction, low advocacy, and diminishing returns from the program. If an advocate genuinely refers a friend, that friend purchases, but you fail to reward the advocate she will likely never share again. You have lost the good will of a good customer. The Extole profile is an advanced tracking and attribution system that matches advocates 30% better than cookie or code tracking. For it to work, we need to be able to evaluate all orders and match based on a range of keys and criteria.

​

Why?

First off, the Extole profile system has superior cross-device tracking and will match more accurately and comprehensively. Traditionally, affiliate programs are tracked by a cookie or a click ID - the prospect clicks on a link and the program would either set a cookie, pass a value through the experience, or distribute a coupon so that there was a local variable that could be used to evaluate a referral. In 2015 Extole worked with a number of clients to analyze historical orders from their order management system, and we discovered that affiliate methods systematically undercounting referrals, which led to higher support calls. More people were being referred and purchasing than were being recognized at the conversion. This was for a reasons that included:

1. The friend opened a referral link on mobile, but then went to the web site on their laptop to purchase because it was easier
2. The friend received a referral link, but they clicked on another link on the email which resulted in their getting to the site
3. The friend clicked and got a coupon, but never used the coupon
4. The friend clicked the link and didn’t buy, but bought a day later after seeing an ad on Instagram

There are likely other scenarios, but they all result in the same outcome when using an affiliate tracking model - a friend is referred and purchases or becomes a new member, but the advocate who referred them is not getting credit when the friend eventually takes that qualifying action. Some marketers may think “That’s great! We got a purchase and didn’t have to pay a reward.” That is the approach behind sweepstakes and rebates, both proven marketing tools. However, advocacy is based on relationships - both the relationship between advocate and friend and the relationship between advocate and brand. When your customer refers and their friend buys but the advocate is not rewarded, they are not happy, and they will not refer again. Moreover, we have seen proof that poor attribution leads directly to expensive customer service issues (we call these WiSMR: “Where is My Reward” calls). Even if you disregard the unhappiness of the customer (which we don’t recommend) you have to consider the real cost of supporting these customers.

​

Performance and Participation

The second reason we don’t support conditional data passing is that our system has auto-promotion features that promote participation by new customers in the program. When you send the email at the time of purchase, we not only use algorithms to match back that purchase to a referral, we also put that new customer into a queue that invites them into the program a certain period after their purchase. This serves to both alert them to the program and to enable active sharing and recommending. This is impossible without tagging.

​

Testing and Targeting

If your initial program is only going to reward if the friend is new-to-file and the order value is over $25, you may consider conditionally sending only those. Those conditions are probably inside engineering code that will take an engineering ticket to change. By putting the conditions inside the Extole program, you can change them or A/B test them without any need to redeploy code. You can also run analysis and forecasting to see how things may change if you lowered it to a $20 minimum.

​

Quality Scoring and Fraud Detection

Finally, we use the total purchase history information to support “new to file” reporting and quality scoring. An essential part of our reward quality engine. It is critical to understand that the Extole software-based profile enables advanced promotion, recruitment, attribution, and segmentation. One of the core integrations to support this architecture is new customers. It makes a better experience, better rewarding, and better participation.

​

Security & Your Data

Because the Extole referral engine requires captures all conversions, we take very seriously the security dimension of our integration. All customer data is encrypted at rest, an advanced standard recognized as essential by large enterprises. All messages are also encrypted in transit. Your engineers may be used to a Facebook pixel, where Facebook owns the data. At Extole, your program is your data. We are a sub-processor and Extole does not use your data for purposes other than running effective advocate marketing campaigns. We do not purchase media using the data nor provide any access to your data accept to members of your client’s team who are there to help you succeed.