Adidas was one of the latest brands to feel this, when its #DareToCreate campaign was subject to abuse of a personalisation mechanic.

The sportswear brand launched a Twitter campaign allowing people to create their own virtual Arsenal jersey emblazoned with their usernames, which was then automatically tweeted and available to order. But the system was unable to identify and eliminate offensive username suggestions. Instead it simply followed the programmed rules, resulting in users creating and tweeting pictures of shirts containing problematic names – a social media disaster for both Adidas and Arsenal.

Of course, this doesn’t mean to say that marketers should resort to attempting to deliver personalisation manually to avoid these situations occurring. But it reaffirms the importance of human supervision and catering to all potential scenarios. With that in mind, we’ve outlined some of the key scenarios where oversights can result in poor consumer experience and most importantly how these can be avoided.


 Very often, returns are not factored into personalisation strategies, resulting in customer profiles that do not mirror the reality and marketing tactics that are unlikely to increase conversion. For example, product recommendations based on past purchases without taking returned items into account can generate irrelevant suggestions as the shopper will have little interest in products similar to the ones they sent back.

Also, as Asos has recently highlighted, a lot of online brands are dealing with serial returners who are buying in bulk before returning most of the items. For marketers, this means that someone may look like a high-value customer because of the volume of purchases they are making, although the majority of these purchases are being sent back and causing significant costs.


Browse and purchase behaviour around gift occasions like Mother’s Day can cause anomalies in the customer data, which can negatively impact personalisation. Not only might the customer receive recommendations and offers based on products they bought for others, but also surprises might be spoiled by retargeting ads if the gift giver and gift recipient share the same computer.

Around major gift giving events, such as Christmas or Valentine’s Day, it is worth tweaking the recommendations strategies and use crowdsourced product suggestions based on real-time trends instead of basing them on the shopper’s purchase and browsing history. Don’t forget to switch recommendation rules back to normal after the events have passed however, so the customer receives targeted communications based on their everyday behaviour.


Whilst retargeting is certainly an effective way to encourage a customer to reconsider an abandoned purchase, there are some exceptions to be aware of. Retargeting a customer with items they have browsed and gone on to purchase isn’t going to drive any additional sales. Additionally, if the item has been discounted since the customer purchased it, retargeting and highlighting the reduced price is going to cause irritation.

In a similar view, recommendations around bigger ticket items and one-time purchases, such as a fridge or a wedding dress, need to be approached carefully. Whilst consumers may spend significant time researching these, and appreciate recommendations to compare products when browsing, once they have completed the purchase they won’t want similar recommendations as these are items they aren’t intending to replace for at least 20 years, if at all. Instead, brands should look to recommend products that compliment the item they purchased as this will be more effective at capturing the shopper’s attention and generating an up-sell.

Ecommerce Best Practice Guide

Pop ups

When deployed correctly, pop ups are hugely effective at encouraging a shopper to share their data with a brand. Used in conjunction with an offer, such as a discount on the first purchase or free delivery, can be the nudge a consumer needs to submit their email address and make their first purchase.

For a consumer who has already bought from the brand, an automated pop up offering 10% off their first purchase is going to have the opposite effect. At the very least, it is wasted marketing space that could have been put to better use, e.g. by inviting the repeat customer to complete a short survey about their recent purchase experience. It could also have a detrimental impact on the relationship with the customer as it indicates that the brand favours new business over loyal customers. Therefore, it’s important that the pop up reflects the stage the customer is at in the funnel to make them effective.

Delivering personalised interactions at the right time and in the right way is critical for driving customer engagement. Whilst automation is critical and very effective at enabling this, marketers should recognise where human intervention is needed to avoid irritating customers. Above all, continually evaluating and reviewing the rules and tactics used in a campaign to stay close to customers is paramount to exceeding expectations and securing repeat purchases.