One of the most prominent technological developments to evolve and radically improve marketing performance in recent times, across email and display advertising in particular, is dynamic targeting and personalisation. It can make a huge difference to engagement levels, conversion rates and customer satisfaction. But just because you’re collecting multi-channel data and using it in your direct marketing materials, that doesn’t mean it’s always going to have a positive impact. It can go spectacularly wrong.

Part of that is down to consumers’ limited understanding of modern advertising personalisation practices. However, whilst the pace of innovation can sometimes be limited by the slower pace of user acceptance and education, the display re-targeting industry is currently experiencing its renaissance. This is, in particular, due to re-formed practices and the fact that consumers have become accustomed to, and begun to see, the value of decent behavioural targeting.

Although this doesn’t apply to everyone. Take conservative politician Gavin Barwell for instance. Earlier this year he tweeted about an inappropriate ad that was being displayed online whilst he was browsing a Labour press release. He was quickly informed by opposition politicians that this ad was actually contextually targeted based on his own browser habits, with embarrassing front-page headline consequences.

Incidents like this, combined with a misuse of the medium by the likes of BT, have naturally established a level of distrust towards ‘big-brother’ labelled practices. So understandably there is a fine line between perceived ‘useful’ value and intrusive content. Once the trust is broken, it is very hard to re-establish. Digital technology affords us the capability to truly personalise marketing messages to the nth degree based on every element of a person’s life and their trackable behaviour.

But it’s clear that there are associated dangers.  One of which is the ease with which companies can be drawn into the public gaze for simple, unplanned side effects that can arise as a result of targeting mishaps. Ad targeting can be incredibly powerful, but you do need to be careful. Contextual use of personalisation, especially across emails, where the legacy creates a permanent trigger and can lead to damage which virally spreads.

So what are some of the common problems marketers can face when it comes to personalisation?

email failPersonalised email campaigns can often be limited or affected by the depth and accuracy of the data that is used to provide the personalisation in the first place. Errors can be caused by incorrect business rules i.e. if Gender = F, then show ‘Dear Mr’ or assumed characteristics where data is missing or broken such as dynamic content tags like HI #FNAME, FIRSTNAME, Unknown, Dear Mr Ash Mark.

Personalisation based on date-specific profile information that is not updated i.e. birthday fields that are static or interests declared in the sign-up process, which are not then revised to reflect changing interests, also throw up unacceptable errors. I’ve stressed the need for testing several times before in my previous articles, but if not reviewed properly, that too can throw up unnecessary errors, particularly when test data is retained in the body of the message.  And when it comes to capitalisation – there’s simply no excuse.

So how do we avoid getting egg on our face?

It’s all about effective quality control. For starters, we need to take precautionary measures when using personalisation to safeguard against inadvertent consequences. For all businesses, this could be something as simple as putting in place a quality control process for your emails. Make sure you check that the data in your database is in the right fields. Don’t take personalisation rules on face value. Using test and approval lists on every email broadcast or senior management sign-off on all brand communications, can significantly help achieve this.

Quality Control needs to replicate every possible personalisation possibility – that means every data personalisation scenario. I had the personal misfortune as a junior marketer of ordering a single excel column instead of the whole sheet which led to the misdirection of an entire direct mailshot to 20,000 addresses. OK, so I was only 20, but I quickly learnt the principle of double and triple checking.

For larger organisations it might mean investment in marketing resource management platforms – especially if you operate in a regulated market. In all cases, having secondary checks and balances in place is absolutely crucial to safeguarding against human error, regardless of how rigorous you think your marketing process is. The dangers for certain sectors such as finance with associated penalties cannot be understated. When it comes to advertising over third-party email lists, you can minimise the risk by buying named site placements and asking email providers for content category exclusions.

Acceptance of targeting and the need for personalisation is now anticipated by consumers, but not too much – the words ‘big brother’ are synonymous with retargeting practices that go beyond any utilitarian value as far as the consumer is concerned and this demonstrates the fine line that needs to be drawn. All this must be balanced by relevance and value to customers within the context of their user journey. Personalisation should also be viewed as a useful function that adds value to your newsletter service. But when it gets misused for the sake of extra opens or clicks, it becomes a bad thing.

A truly personal email addresses the user’s needs, desires, fears, preferences and expectations. A lot of this falls more under the banner of segmentation, especially when we think about it from the subscriber’s perspective and as such it should therefore be seen in that context. Don’t just personalise for the sake of it. It will come across as lazy, automated and rather insincere as a consequence. You don’t want to undermine yourself either by sending unconnected follow-up content.

Consumers’ experience of personalisation is often affected by the relevance of the content that is being personalised. This might be arbitrarily created based on their initial interest but interest will almost certainly vary over time. Opt-in is a perceived value (regardless of what someone ticks in a box). It is based on relevance, engaging messaging and acceptance. It’s not a one-time event. Inbox deliverability as we know is more affected by user interaction than ever before.  Make sure that users are aware of content/offers that are personalised and invite them to change their preferences on a regular basis or comment on the different elements of your email content. Not only does this allow users to control and personalise their own content so that it is more relevant to them, it also reduces unsubscribe requests and provides you with greater insight into the needs of your customers.

Testing is crucial to understanding people’s acceptance and interest sweet spots. Technology allows you the opportunity to find the optimal levels of targeting and personalisation – so make sure you use it. Test every perceivable data variable to make sure that business rules are running as you expect them to – especially nested rules which can behave incoherently if they are nested within campaign level filtering rules. If it’s impractical to look at every personalisation outcome, run filters on the data to ensure that the segmentation rules you have provided roughly produce the counts you were expecting. If nothing else sanity-check the data and export it to make sure it is in the format you anticipate. If you do, you’ll be able to deliver the level of personalisation you expect and reap the associated benefits.

Mark Ash

Mark Ash


Mark Ash is Managing Director Teradata Interactive International.