Cookies are shrouded in mystery and distrusted by many. But are they the figurative ‘bogeyman of the 21st century for websites and their visitors’ or merely misunderstood?
When used and configured properly, cookies can provide us with an unparalleled user experience. We want to debunk the myths surrounding cookies, explaining what they are and how they work, to answer the pressing question – what do your cookies say about you?
What is a cookie?
A cookie is a small amount of data that is generated by a website and saved by your web browser. In basic terms, its purpose is to remember information about you. A cookie’s most common purpose is to store login information for a specific site. It can contain information such as your preferred language, session tokens, user preferences, and everything else a website needs to track a visit.
How do they work?
When you visit a website, a request from the website to the browser is triggered to set the cookie. The next time you visit a page it sends that cookie back to the website. Each request is totally separate from the following one, so the server needs an approach to monitor which request belongs to which visitor. By storing a small bit of information in a cookie, the site can establish that the online visit belongs to an individual user account. Without cookies, surfing the web would be a much more frustrating experience.
What’s in a cookie?
The best way to understand what is in a cookie is to analyse it. So as an experiment, I looked up the information stored about me via a data information company that allows consumers to view the 3rd party interest data related to their individual computer and browser.
The site can show consumers information relating to their perceived location, professional interests, purchasing history and more, with the aim of encouraging transparency between user and advertiser.
I’ve looked at the data stored about me on my personal laptop and browser to see just how accurate – or inaccurate - it is.
1. My basic info
• Home owner
• No children
• Current student
• Undergraduate degree
• Stocks and bonds
According to my basic info data, I am a home owner, have no children, am a current student, hold an undergraduate degree, am both male and female, am unmarried and have looked into stocks and bonds.
Is this accurate? Whilst a lot of the above is true, I do not own my own home and am not a current student or male. If I were to be served an ad intended for a male home-owner, for example, I would be hugely unlikely to engage and could even be frustrated at the advertiser. This is a classic example of bad audience segmentation and shows how unreliable 3rd party sources can be.
2. My location and neighbourhood
• Personal care
• United Kingdom
There is little information about where I am based in my location cookies, though the data accurately asserts that I am UK-based, which is encouraging. However, it is surprising that there is no cookie that relates to me living in London, which would be a more granular data point for advertisers to target. It’s hugely important for advertisers to know the location of their prospective user for lots of reasons, to ensure ads are served in the correct language or not advertise items that are unable to ship, as examples. The Personal care cookie relates in some way to healthcare and beauty, but there is no further explanation on this data point for me to analyse.
3. My professional interests
• Information technology and computing
• B2B marketing
I work as a Content Manager in the eCommerce and technology industries, so the three professional interests associated with me are impressively spot on and relate to my everyday work.
4. My hobbies and interests
• Luxury cars
• Foodie and gourmet
• Personal savings and investments
• Online news sites
• Performing arts and live theater
• Business and finance
There are a whole twenty pages of points that relate to my hobbies and interests, but for brevity, I’ve shown a sample of data points above. Of the first twenty-five cookies, I would count fourteen as my legitimate hobbies and interests, with eleven being completely unrelated to my life. It’s especially surprising to see cars associated with me when I have no interest in the automotive sector and don’t own or want a car. This is quite worrying information as advertisers could be spending time, money and resource serving ads to me that I have no interest in.
Ve’s Display Analytics Director, Sam Baker, explains that,
"3rd party data can be completely unreliable when it is invalidated in this way. Companies have got to know that the data they use is being processed properly before they rely on it to target users."
5. Things I might want to buy
• Clothing, shoes and accessories
• Video games
• Sports equipment and outdoor gear
I do a lot of varied online shopping, so I was surprised to see so few data points related to things I might want to buy. However, the limited number of cookies could mean better qualified, filtered data that is based on related purchase intent, meaning it is more accurate. Of the seven cookies listed above, though, the two that don’t apply are autos and HDTVs. I don’t own or want a car and I already have a TV so these two data points are imprecise. Five out of the seven points legitimately refer to things I might want to buy.
6. Things I might have bought
• Sauces, spices and seasonings
• Women’s retail items
• Used cars
• Auto insurance
• iOS (iPad)
• Desktops and laptops
The majority of the data points are correct, though I have never researched or considered auto insurance or used cars. Some of the first twenty-five points related to me also seemed a bit strange. But after looking at past purchases, the data started to make more sense. ‘Sauces, spices and seasonings’, for example, appeared to be inaccurate and random. However, I analysed my shopping history and discovered that around six months ago I bought a spice mix box as a present online.
Yet having this cookie is not necessarily a good thing. Accessing historical data related to user behaviour can be really useful for companies. For example, a customer who buys a pram online, might be interested in clothes for toddlers a year later. This knowledge of previous online behavior could help advertisers dictate future ad serving and assist in providing excellent customer service. But with my example of the spice mix box, it was a one-off purchase that I would not consider buying again. So if I get served spice-related ads, I become a frustrated user.
It’s a dilemma, and one that Sam connects to recency,
"When it comes to managing data, it’s never as simple as having a cookie for life. The recency has to be accurate and users have to be treated as active buyers, not data points. Analysis of purchase patterns as well as thorough cookie adaptation and maintenance are pivotal to inspiring repeat purchases."
7. What others know about me
• Health and fitness
• Books and magazines
• Electronics and gadgets
• Social media users
• Government and politics
• News and current events
• College degree
There is a wealth of information about me in this data section, which makes sense as it is pooled from a mix of third party sources that have tracked my online behaviour. Though as Ve’s Chief Product Officer, Jean-Baptiste Theard, asserts,
"It is not so much about big data – it’s far more about intelligent data and what you can do with it."
I’m impressed that the first five pages of my cookie data are, for the most part, accurate, although I have no interest in or shopping history related to basketball or automobiles.
The sheer amount of data about me in this section (over 500 cookies) might seem like Christmas come early for advertisers, but Sam warns against the deluge of data that as an industry we now have access to:
"There is so much data available to advertisers out there right now, and this is only going to increase with time. But it doesn’t matter how much data you have if it is not processed properly. And it’s how you target that ultimately makes all the difference. ‘Everyone has access to hordes of third party data – but is it accurate?’ This is the question that all advertisers need to ask."
So what do my cookies say about me?
And do the data points build an accurate persona of me as an online shopper and browser?
In general terms, the cookie data matched my online behaviour and shopping persona. There were, however, some key discrepancies. I do not own my own home and am not a current student or male. I also have no interest in or known history related to basketball or automobiles. These pieces of information are not just slightly ‘off’, they are actually wrong. This is quite worrying, as companies who are using this data to target me will not be able to personalise effectively with content I would be interested in.
Here at Ve, personalisation is key to the success of our solutions. Because of this, we build our own segments, which we qualify over time, so we know first-hand what makes up our audience cookies.
This enables us to create an in-depth and accurate understanding of the users we interact with, allowing us to know who, where and how best to talk to the users we advertise to, based on the most relevant cookies.
The data we model is from publisher partnerships, many of which are unique to us, and opted-in advertisers. Sam explains further,
“There’s a lot of fear and confusion surrounding cookies, but ultimately they’re a good thing for online users. At Ve we have unrivalled data processing ability, that allows us to target based on in-depth intent signals. Where one advertiser might target a user because they visited a site just once, for example, we look at time on page, number of visits, onsite activity and more, to determine who that user is and what they as an individual want to see. We champion ethical targeting that benefits both user and advertiser.”
So what do your cookies say about you? My research shows that it’s a real mix of relevance and inaccuracy, illustrating the need for validating audiences. At Ve, this is at the forefront of our digital strategy, meaning minimised wastage and accurate targeting.
Get started with us now to learn how you can target your users better.