Web Personalisation is more complicated than you thought
Website visitors do not want obvious personalisation as much as they would like to believe something is personal
The chances are you have received an email addressed to “Dear [First Name]” or similar? Email marketers appear to forget to include our actual names from time to time. It is annoying and we all laugh at it, realising that the company involved has failed to populate its database properly or set up its email software correctly. However, do we care? Are we bothered if the personalisation of emails is poorly carried out?
Ask professional marketers and there’s a good chance that they’ll tell you that personalisation is essential. Indeed, according to a report from Monetate, 94% of businesses believe that proper personalisation is “critical” to success. Meanwhile, less than one in every six firms felt that they were getting a return on their investment in personalisation.
Perhaps that is because people are not as bothered about personalisation as many marketers think. New research suggests that what is important is “perceived personalisation” rather than actually being personal.
Names do not matter much
The study conducted at the University of Miami included an interesting experiment in which people were addressed either by their real name or as “Dear Consumer”. The level of interest in what was being sold was then measured. There was no difference in the degree to which people were prepared to consider what was on sale. Whether they were addressed by their real name or as “consumer” made no difference to their level of interest.
This goes against the long-held theory that names matter and that “Dear First Name” is a negative turn-off. What appears to matter most is getting it wrong. The Miami research found that when they called someone by the wrong name, that led to disinterest. When someone who was called David was addressed as Kevin, for instance, they became completely disinterested in what was on offer.
What this research shows is that we are not that bothered about being called by name. The study included a third experiment that shows why. The preferred holiday destination of participants was known in advance. These people were then asked to look at a web page where a holiday advert for the kind of destination was included. For some participants, the name of their favourite destination was included in the advert. However, it transpired that people were equally interested in the advertisement, regardless of whether or not it contained the destination name. It appears that the mere notion of the holiday was enough to convince people that they should be interested in the advert. The presence of the name made no difference.
Target people’s interests
The three Miami experiments show that as long as people think information is personalised to them they are interested – even if there is no actual personalisation. The “Dear First Name” error is not a problem if the remainder of the email contains information that is of interest to the recipient. Because of this, they think the material is personalised.
Of course, this assumes you know the real interests of your target customers. This is where web technologies let marketers down. Web page advertising, for instance, uses past browser history to target adverts to people. However, it becomes incredibly annoying when you always see advertisements for items you have already purchased. No longer is the advertising about something in which you are interested. As a result, it is like calling David by the name Kevin.
Similarly, many e-commerce sites use visitor purchase history to present a page of options that is “personalised”. However, this frequently fails. If you have been shopping for gifts for a friend at an online retailer you are not that interested in the items yourself. Hence, the so-called “personalised” page is nothing of the sort. It would be better suited to your friend.
For marketers, there is another problem. People do not like their activities being tracked. In one study conducted by Adobe more than two-thirds of people say it is “creepy” when websites track their activities. Increasingly, too, web browsers offer private browsing and Internet security suites are preventing the data from being collected in the first place. The result, inevitably, is going to be that just as marketers start to be able to gather more information to provide greater levels of web personalisation, users are going to stop that from happening.
How can you collect useable data?
People want apparently personalised material and they like web pages that are targeted to their precise needs. However, the desire for privacy is making this harder to achieve. So how can you make people believe the web page they are looking at is personalised to them? After all, the research shows it does not need to be personalised, but it needs to appear to be personalised.
One way out of this is to not personalise at all but to provide people with choices, allowing them to “drill down” to what they are interested in. Human beings are really poor at making decisions. When we are faced with more than three things to select, we end up with “analysis paralysis” making it almost impossible for us to decide what to do.
However, you can make people think the page is personalised without doing any personalisation if you let people choose things from up to three options. Imagine, for instance, you want to market software. Visitors could land on a page which asks them to select “Windows, Apple, or Mobile”. Let’s assume they select “Windows”. That leads to three more choices, such as “Office, Utilities or Productivity”. If they click on “Productivity” the website then offers them “Time Management, Note-taking, or Filing”. And so on. After four clicks they are on a page that has precisely what they are interested in; it has the perception of being personalised.
To achieve this, you have not needed to collect any personal data. Neither have you needed any fancy web scripts to produce dynamic pages.
However, for large e-commerce sites, such a system would require too many clicks and hence you would need an additional way of achieving apparent personalisation. One way of achieving this would be allowing people to answer a short questionnaire or poll that provided data to produce pages of interest. Indeed, one way of personalising things without doing “creepy” data collection from browser activity is simply to ask people what they want.
Talk to your customers more
One of the benefits of web technology is that we can collect so much data from our clients and targets. However, that has meant that many businesses now rely more on that data than they do on actually speaking with people.
If you consider a bricks and mortar retailer, staff are trained to talk to customers and to personalise the shopping experience. However, they do not usually ask for our names or any personal details. All they do is ask a couple of questions that elicit the kind of thing we want to buy and the reason for purchasing it. Then, the sales assistant focuses their information on what we are interested in buying. To us, it seems like it is personal, but to the sales assistant it is just another customer who is told information that is about the stuff they are looking for. It is apparent, not real, personalisation and it has worked exceedingly well for retailers for several hundred years.
However, how do they achieve this degree of perceived personalisation? They talk to their customers. That is all. No fancy data collection. Just a conversation.
Online, many businesses have started too much to rely on data collection to produce real personalisation. However, as the research shows real personalisation is not what matters – apparent personalisation is more important. You can achieve that by talking to your customers – or asking them questions when they land on your website. There is a role here for webchat systems too, clearly.
Maybe those 94% of businesses looking to ensure proper personalisation are looking in the wrong place. After all, most of them are failing to make returns on their investment in personalisation technologies. Perhaps they could spend less and make higher profits if they did not strive for real personalisation but produced perceived personalisation instead, partly as a result of talking to their customers more frequently.
[Editor’s note: Thanks to Graham for highlighting the challenges of personalisation and how to get it right. For more detail on best practices for implementing Personalisation, Expert members can follow the recommendations and examples in 4.2 Web Personalisation in our RACE Digital Marketing Elearning course].
From our sponsors: Web Personalisation is more complicated than you thought