Liar’s data: Is your data telling the truth?
With so many people lying about their age, income and weight online, how accurate and valuable is the data marketers get to work with, asks Tyler Greer?
For almost the entirety of the 1980s, my mother was aged 32. At some point she decided this was the number she was most comfortable with to describe who she was, how she felt, and what she thought she could get away with.
She was neither the first nor last to cook the age books, and we often take it for granted that people will take the odd liberty when describing themselves. And whilst that may be comforting for the person making these claims, it makes it problematic for those using this information in any critical planning.
‘I love you’. ’The cheque’s in the post’. ‘This data is accurate and free from bias’. The latter might not be quite as well known as the first two famous lies but if you work in marketing and you are reliant on purchased, third-party data then you should bear in mind that people lie – particularly when they are asked to describe themselves.
This means that, like my mother, they often fudge the most basic of demographics such as age, gender and income.
Consequently, if you rely on buying data then it’s a case of ‘buyer beware’. Some sources need to be approached with a healthy degree of skepticism because they are the playgrounds of those who are most elastic with the truth. To give you a heads-up, here at Exponential we’ve crunched the numbers and come up with a list of usual suspects:
Love me long time
Probably the most obvious home for ‘creative’ description and therefore skewed data, but most of the free dating sites make the lion’s share of revenue from selling information to advertisers.
I’m married, but from conversations with my single friends, it won’t come as any surprise that a recent University of Wisconsin study found that 81% of people who are dating online lie about at least one physical attribute (height, weight or age).
Given these are all factors that an outside party could notice right away then the rest of the data looks even sketchier (and perhaps the reason for the Massachusetts police blotter entry about a woman feigning illness on a hiking date to get out of it that did the rounds in social media a few weeks ago). If that many people are lying about something easily observed, how many are willing to lie about things that are not so obvious?
Unicorns have long noses
To get ahead in business then you need a lot of chutzpah so it’s not a huge stretch to find that entrepreneurship-oriented sites are knee deep in it online. Exponential did its own research into these sites and whilst it was US-focused, we expected to see the results track in Australia.
According to our deep-dive into anonymised users, even personal income data from self-styled entrepreneurs is significantly inflated.
For example, users on venture capital sites are 8.55 times more likely to over-report their income than the average internet user, closely followed by funding sites (8.51 times), franchise sites (7.73 times), and start-up content (7.42 times).
Check their references
Another place where users tend to overstate their salary is on career sites. Income data from users on these sites is 6.28 times more likely to be inflated. Unlike our friends on dating sites who lie about obvious physical characteristics that will be found out immediately on meeting in real life, this lie at least makes sense.
Users on these sites are presenting an ideal professional persona to make them more attractive to head hunters — and more worthy of a higher salary at their next gig. But this isn’t good news for advertisers who gather income data on these sites (and also a reminder it’s imperative to follow up on references).
Indebted
On the flip side of the coin, there are sites where people underestimate their income – namely loan management sites. According to our data, people are 8.56 times more likely to under-report income on student loan sites and 9.75 times more on general debt sites.
Debt can be a huge burden and so it scrambles an individual’s perception about all financial matters.
Good quality data is one of the internet’s greatest gifts to marketers. When people consume content online they send signals, allowing us (the marketers) to better understand who they are and what they are interested in. Simply put, how people act online often tells us who they are offline.
But to properly take advantage of this golden data it is critical to understand its provenance. How is the data gathered, how is it collated, how granular is it, and how old is it? Each question is critical in understanding its value.
And most importantly, add intelligent analysis to good quality data. After all, human nature is such that we can’t prevent people from lying online (or elsewhere), understanding the psychology of the ‘why’ they lie is half the battle.
For example, if consumers on loan management sites lie because they feel burdened by debt then that is your consumer truth and there are some brands that may be able to market products and services that address this anxiety.
Approach data with eyes wide open and apply context and insight. Perhaps then even the biases can turn out to be your friend.
Tyler Greer is head of strategy, APAC, at Exponential
Interesting read. How did you uncover real incomes for benchmarking? Definitely agree that understanding the psychology of the ‘why’ is the foot in the door.
User ID not verified.
need to do analysis on asset maturity data to show how to grow asset
User ID not verified.
Good points, well made Tyler. Tid 32-year old male has been a 70-year old Woman online for as long as I can remember.
The stream of menopause, cruising and financial planning ads make it all worthwhile
User ID not verified.
*This
User ID not verified.