Twitter Knows What You Did Last Summer

Twitter knows everything about you. Sort of. If twitter knows as much about you as it does about me it’s very likely got a few things right, and a bunch of things wrong.

Twitter thinks I’m interested in Ava DuVernay (I’m a fan), books and literature, which matches some of my personal interests. And then Big Data, Brands, Digital, Leadership, Travel, Books, Technology, so far so good!

But Twitter also got a lot wrong – it thinks I’m also interested in Automotive news (I’ve never owned a car), Baseball and MLB – um no, Soccer, Coca-cola, Fashion (barely), Office 365 and Uber. Also a whole bunch of presumably famous people that I would have to look up to find out why I am interested in them, I’m a bit hopeless on famous people. The whole thing is a bit like reading your horoscope, 25% me going “oh yes, that’s me” and 75% me going “so much no”.

You can check what twitter knows about you by going to “Settings and privacy” > “Your Twitter data” > “Interest and ads data.”

I’ve turned off the option to share my interests with partners because I’m a bit paranoid about data sharing and Twitter already annoys me with their notifications. But it doesn’t seem to go well, Vox reporter Emily Stewart shows the interests that came up for her, including family status, salary estimate and gender.

How does Twitter figure out these interests – I think mainly via tracking cookies that are picking up on my search terms, I admit I may have searched a few football/soccer related articles.

What does twitter do with them? Apparently throws that data into an algorithm to deduce the demographics to create those groups that interest advertisers. If that freaks you out, you can opt out of it.

The big platforms, including Twitter, tell us that more relevant ads will be Good For Us, in fact they’re good for the advertisers in the sense that sales rise. Harvard researchers found that when the viewer was aware of the techniques used to increased the relevancy of those ads, and judged the methods as “creepy” ad effectiveness declined. It’s a reaction I’ve noted in myself, and probably what led me to stop following brands on Facebook and to turn off the ad customisation options on Twitter in the first place.

Image by StockSnap from Pixabay

Just Stop It – Don’t talk to me in a funny language

Just Stop it

Last Month YouTube updated their terms of service, and users in the EU and Switzerland had to agree to new terms so I got this notification.


So I clicked on it, and got a document of more than 4,000 words in Dutch.

I can read Dutch, but it’s much slower for me than reading English. I know YouTube must have this document in English because Ireland and the UK are (at date of writing) part of the EU and YouTube is an American entity so it’s highly likely that they created the document in English. So I’m sure they have the right content – and a quick search revealed they do.

I use the internet in English at least 90% of the time, my language settings are all for UK English, my browser is in English, my YouTube account specifies UK English.

But it seems that YouTube have chosen to use my IP address to determine which language I get my terms and conditions in. This is a Bad Idea, it’s a very poor data point to predict language.

  • Internet access can be routed through another country, my work computer can go via UK, Netherlands, Singapore or the US depending on which data centre I route it through.
  • People travel, within a 2 hour flight I could be in a country where people speak French, English, Irish, Danish, Swedish, Italian, Spanish, German, Czech, or Polish.
  • In some countries there are multiple languages spoken, what did YouTube do to the Belgians?

Websites can pick up the language of the browser, that’s a better guess at which language to deliver content. And in this case I was logged in. I TOLD YouTube what language I wanted.

American companies are really bad at this, they need to hire more Europeans to their UX teams. Hire some Belgians, YouTube.

Cookie Nightmare

CM2016_09_cookie.png

Do you know how many cookies are placed on your computer? Does it matter?

The EU directive from 2011 had companies scrambling to find good ways of notifying visitors about the cookies being placed on their computer and giving opt-out measures. There weren’t good tools around and translating the law into technical requirements was a bit of a nightmare. Ironically it led to the company I worked for collecting more information, as we needed to be able show that we’d responded to people’s cookie preferences.

There are three common approaches;

  1. Implicit agreement
    a warning is placed on a website saying that if you proceed with viewing the website you accept cookies from the publisher, this is most common on information or news sites, it seems to be more common on UK sites than Dutch sites, here’s how the Guardian presents their cookie notification, they also offer a detailed explanation of cookies.
    Screen Shot 2016-09-06 at 09.04.38
  2. Forced agreement the site is blurred out or obscured and an overlay forces you to click ‘agree’ to proceed, this is commonly used on Dutch sites, here’s the Dutch newspaper Het Parool, you only have the option to accept.Screen Shot 2016-09-06 at 09.00.53
  3. Pop-up with cookie options
    This is rarer, but some sites give you the option to choose which cookies you would like to install, usually giving you a choice of three levels of cookies. The lowest level is those required for website function, the next level relates to site measurement or personalisation, and the third level is often the advertising cookies. It’s this third level that mean you’ll see ads from the same company every time you open the internet for 30 days, no matter which page you’re on. The advertiser is collecting significant information about your site visits.

I’ve heard from web experts that the number of people adjusting the level of cookies they accept is low, less than 1%, which makes it seem a lot of work to manage cookies for a very small group of people.

However Many people manage their cookies though browser settings, it’s fairly easy to do in Chrome and Firefox,  and I suspect people really concerned about cookies and privacy take such measures.

When the ‘pop up with cookie options’ is used it’s not always clear how to find the cookie options. One of the most common tools used by companies (who often outsource the cookie management) is TRUSTe, which does give visitors control of their cookies but it’s not easy to see how.

When opening a website using TRUSTe you are presented with a pop-up that talks about “Your Choices” but is designed to push you to clicking on “agree and proceed”.  The little link to the right, that doesn’t look like it does anything is actually where you find the choices.

cookies1Here are the three choices you’ll get.

cookies2Required cookies just let the site function in a sensible way, it means the site will “remember” your language preference for example, sometimes the cookie only lasts for the duration of your visit. Functional cookies provide data on your visit and advertising cookies mean your data is going to an advertiser or media buyer – these are the cookies about which there should be the most privacy concerns.

In all the cases I’ve checked the default setting is for advertising cookies.

I changed the setting to allow only required cookies, and got a warning that the submission would take up to a few minutes.

cookies3

In fact it took less than a minute – this time.

I think some cookies, like those retaining a language preference, on-site tracking or login details, do not cause any significant privacy issues. Others, the advertising cookies, the tracking cookies, are a potential issue. Yet, despite all the good intentions of the EU directive, only one of the cookie options implemented allows you to opt out of those cookies and that’s not always easy to find.

How do you manage cookies as a visitor? I’ve put a poll up on twitter, let me know on the poll, on twitter, or here in the comments.

 

 

Header Image; Halloween Sugar Cookies  |  Annie  |  CC BY-NC-ND 2.0

Easter Egg Pheneomenon

The ones found in media, the chocolate ones are for the weekend.

An Easter Egg is a surprise addition, something unexpected and usually humorous, included on a DVD, movie, music compilation or software. They usually have an “inside joke” quality to them, and some range into rather esoteric geek territory.

Online Easter Eggs

Google leads the way in producing browser-based Easter eggs with easy to get jokes. They’re your standard hollow chocolate Easter egg. Easy enough to consume, and leave you wanting more. Here are some of my favourites;

Search for ‘anagram’, ‘recursion’, ‘askew’ or ‘do a barrel roll’ and watch what happens.

If you use google maps, pay attention to what happens to the pegman, he changes in various locations or to celebrate specific occasions. He’s been a penguin, a witch, a leprechaun a rainbow, a skier and an astronaut. There are two locations I found that still have adapted pegmen.

 

This slideshow requires JavaScript.

There are also some hidden subpages from Google, a virtual teapot for example, or a bonus puppy shot in the app store.

Easter Eggs In Films and TV

Disney is famous for “cameo” appearances of one character into another movie, so Goofy turns up in the Little Mermaid, and Mickey Mouse makes a brief appearance in Frozen.

The geek TV sitcom “The Big Bang Theory” has a final shot that’s a “card” from Chuck Lorre, the series creator. These are usually on the screen for a matter of seconds meaning it takes some dedication to get to read them. Card #221 is a philosophical rambling on politics and horses.

Easter Eggs In Advertising

Ever wondered why phone numbers in (US made) movies begin with “555”? It’s a range of fictitious phone numbers so that the movie doesn’t accidentally use a working phone number. Except when it does. In 2014 Old Spice included a real phone number in their ad and gave away free super bowl tickets to the first person who called the number.

More delightfully, Innocent, the UK health drink company, included a help line phone number on their bottles. They invited you to call them even if  you did not have a problem and promised to sing you a song if you did. We called and they did indeed sing.

Easter Eggs In Geekdom

I’ll stay out of the seriously geek territory, but will point out that if you open a firefox browswer and type ‘about:mozilla’ into the URL bar, you will get to read a verse from the Book of Mozilla. The verses are written in an eerily apocalyptic style, but do contain references to events that are history for Mozilla. With some historic knowledge you can decode them.

I think companies, such as Google and Disney, which create Easter eggs in their products are displaying a sense of fun – sometimes to the surprise of unsuspecting viewers. They’re also inviting us in, teasing us to become part of the inner circle. Not every company has the platform to do this, nor the company culture to support it, however when it works it’s genius branding.

Still looking for the perfect chocolate Easter egg?  Michel Roux jr reviewed various flavours, and then there’s this, a Game of Thrones inspired dragon’s egg.

Dragon's egg

 

Image: Easter eggs  |  Dan Zen  |  CC BY 2.0

 

The Internet of Things

The idea behind the Internet of Things (IoT) is simple, use the internet as a communications network and enable devices to talk to each other, to applications and ultimately to us.

To give a really simple example that already exists, in fact I have one installed in my house, a home heating system that includes a thermostat and an app that lets me schedule temperature changed by date and time.  It cost an extra hundred euros, but the savings due to scheduling lower temperatures at night and on holiday will pay for that.

IoT has tremendous potential to simplify our homes, cities, work environments, transport systems, and healthcare. We’re well on the path towards the IoT, many companies are creating connected devices and Intel has stated that we’ll all have 7 connected devices by 2020.

ISF_Infographic_1600x944

Most people already have multiple devices that are connected to the internet; phones, laptops, e-readers, computers, TVs. Coming soon to your home are connected devices such as heating, fridges, lighting, sound systems, and home security. Gartner estimates that by 2022 we’ll have more than 500 devices in our homes, although they don’t provide a list.

IoT goes beyond our front door, the healthcare industry is looking at connected devices to support patient care in hospitals and live-at-home independence for people with disabilities and the elderly.  For those without known health issues wearable devices monitor your activity and fitness each day.

1970s Teasmade

Many of the home devices can also make our work environment smarter, and more productive, including a wifi enabled coffee machine reminiscent of the old fashioned teasmade of the 70s.

IoT also impacts our transport systems, giving us the famous Google driver-less cars and smart driving systems, along with automation of our public transport systems.

On a large scale cities are looking at smart ways to use limited resources, including space, more effectively. Monitoring traffic, pollution, rainfall, foot traffic, waste disposal can help a city provide better services and save money. A very simple example; lighting a city can be revolutionised by knowing how city space is used, and the lights themselves can be monitored and maintained based on information rather than inefficient repeated inspections.

What’s the catch?

Masses of opportunities, but what’s the catch? There are concerns around security, privacy, and some specific ethical concerns.

If your connected device is critical then it needs to be secure, hackers have already tested a number of devices and found that the security is lacking. In one alarming case researchers hacked a pacemaker, the pacemaker was in a mannequin, but if it had been in a person that would have amounted to a death sentence. Some guidelines have already been created to protect yourself against IoT risks.

If our homes have hundreds of connected devices how can we know which data is provided? Many of the IoT devices don’t allow you to discover that. There are existing data protection laws in place that companies must follow, but when each “thing” in your portfolio of IoT is transmitting data about one aspect of your life that is a massive amount of data.

Driverless cars, potentially part of the IoT pose a very specific ethical challenge; how should they be programmed when the choice is between harming a passenger vs harming a pedestrian? I don’t know either – and the dilemma is likely to push us towards smart assisted drivers rather than fully driverless cars in the short term.

I’m excited by much of this development, but if devices remain discrete and unconnected the number of control apps I have on my phone will become unmanageable, this is starting to be addressed with some platform systems for smart homes. I can’t help wondering what I will do with all this new information, and whether it will really give me new insights.

Image:  BB8 via pixabay

Designed to tell the company story

Luxottica, known for its product design, has just updated its website to reflect online design trends – and give its products a stylish showcase.

Luxottica’s site last week – showing the home page and the brand page.

The site served the purpose of communicating company performance but did little to inspire interest in design or products. Given that they design eye-wear for some of the world’s great fashion brands it was a disappointing experience.

Luxottica’s site now – showing the home page and the brand page.

The difference is huge; Luxottica has re-used a lot of existing content to create a rich experience for the user packed with images and video. The navigation is simplified,  brands are highlighted, and the company’s charitable foundation “OneSight” is featured.

But the changes go deeper than just visual, they include;

  • responsive design, meaning this site will look good on all devices
  • shareable content, every page includes the “share” option under an icon
  • pulling in content from social media channels
  • icons used to identify functions across the site
  • “infinite” scroll, combined with persistent left hand navigation
  • increased storytelling, instead of writing text about the company or the brand stories have been collated from across the company to give the visitor a understanding of the whole company.

This website design is on trend, covering a number of the 7 digital design trends I wrote about last year.

There are a few “minus points”; the media gallery includes just 5 images which seems very thin when the rest of the site is so rich, the videos are sometimes very long,  and the content in the individual brand pages is rather uneven (very rich for Oliver People and very thin for Chanel) which I understand is due to some brands being owned while others are licencing agreements.

But the framework is there to deliver great visual content, and tell the brand story to all stakeholders. The team behind the site should be congratulated, it’s a great step in the right direction – Luxottica.com now looks like it’s from a design company.

(Disclaimer; I know the project manager behind this, she’s fantastic – she also used to work for me)

Increasing Web Traffic is not a Business Goal

Imagine you’re the director of a fantastic, imaginative and popular theme park, located 30 kilometres from a European capital.

How would you measure the success of the park?

  • by the number of visitors to the city
  • the number of tickets booked online
  • the number of people through the park’s gate
  • total revenue
  • revenue against costs

If you chose the fifth option then you can probably stop reading.  While all the others measure factors impacting the business it is the only measure of the success of the park.

Imagine the you ran a campaign to increase visitors and doubled the number of people coming through the gate; but they came on discounted tickets, didn’t spend as much once in the park, but drove up service costs. If your KPI was only on gate numbers you’ll think this was a success, but on a business basis you’ve destroyed value. And in the long term these may not be the loyal customers you’re looking for (as many Groupon suppliers found to their cost).

This is a rough analogy of measuring web traffic, it’s a contributing factor to business success, but not an outright measure of success. It’s also something you don’t entirely control. Sure you can do all the SEO and banner campaigns to drive traffic but external factors also play a part; the biggest traffic drivers to our corporate site in recent years have been events around the financial crisis.

So if you’re trying to develop a set of KPI’s start with the business goal, which should relate to either increasing revenue, building your brand (which should lead to increasing revenue) or reducing/optimising costs. Look at the contributing factors, understand their impact. If you’re looking at website traffic analyse the data in depth, try to find the behaviours that contribute to your business goals. Is it a sale? a subscription? sharing content? viewing a video? Measure that. Measure the number of people who do that as a proportion of total visitors. That’s your conversion rate, that’s the interesting number. To go back to the theme park analogy those are the people signing for the all-inclusive deal.

Traffic isn’t the only thing to think about. Some years ago a Google sales person was talking to me about increasing traffic to our corporate site. At the time my concern was that we had too much traffic – because the site uses the .com domain US clients often expected it to be their “local” site. So it’s worth using surveys to analyse who is visiting your site and what their goals are – in our case we address this specific issue using IP sniffing to guess the visitor’s location, and then served them a splash page directing them to the local site (since some US visitors do want the corporate site we couldn’t just redirect US traffic). So it’s not just volume, it’s whether you’re bringing the right people to the site.

Traffic to a site or within a site should be measured, and web managers must make adjustments  that make their site easier and faster to use. Increasing traffic will almost always be good for business – just don’t mistake it for a measure of business value.

Image traffic

Confession; I was a troll

In my (internet) youth I was a troll, I did it for fun.

I’d pick a cryptic handle such as birds_of_paradox, and tease the other regulars on a forum. I once posed as Mrs_God to counter-troll an unpleasantly bossy Mr_God. I wasn’t ever nasty or abusive, OK maybe occasionally a low grade of mean. So for me it was adopting an anonymous handle and teasing a bunch of people. I’ve moved on, most of the stuff I post on the internet now is in my own name.

Of course I’ve encountered more sinister forms of trolls; some were simply out to challenge political views, some would play devil’s advocate against whatever the discussion was, some posted irrational statements to draw attention to themselves, some posted porn images deceptive titles, and there was the inevitable Rickrolling. I’ve even had one troll want to meet me. Er, no thanks.

But those were the good old days. Trolling seems to have gone high octane, with certain twitter posts including threats to a celebrity or their family, an Olympic athlete was abused via twitter last year, and this year’s Women’s Wimbledon champion was abused for not looking like Sharapova (interestingly a number of those twitterers have now locked or closed their account). The people doing this are using the seeming anonymity of the internet to abuse someone who doesn’t deserve it… seriously; who abuses someone for winning a grand slam tournament?

The meaning of the word has shifted in a second way, it’s not longer about teasing regulars on a forum, or challenging a collective viewpoint, or even getting an angry reaction from a message board. It seems to be used to describe anything that annoys someone somewhere on the internet, including something published by a mainstream news organisation such as the Rolling Stone’s cover photo of Dzhokhar Tsarnaev. As Salon quoted in a recent article;

“People have come to use the word ‘troll’ to mean, ‘It made me angry on the Internet,’” said Doyle. “And that’s pretty broad. It’s a big and noisy Internet.”

The meaning of words changes over time; “nice” used to mean stupid for example (and now you’ll be suspicious if I ever use it to compliment you).

But the problem here is we already have some words that work; in the first case how about “abuse” or “bully”. In the second “provocative”. It was a provocative cover, designed to provoke a reaction or challenge perceptions.

Meanwhile I’ve got some new hobbies – I’ve abandoned my troll bridge.

Any other reformed trolls out there?

Image; Troll 

The Law of Cookies

About a year ago the EU directive on the use of cookies online came into effect. The idea was to give consumers a better understanding of what information was being collected about them and how it was being used. Which seems like a noble motivation. Many experts said that the law was unworkable and ineffective – even as they scrambled to implement it on their sites. Others pointed out that it was unenforceable outside the Netherlands.

The Dutch requirements are, it seems the toughest, and implementing them means we now collect more information than we used to, and store it longer. Because the law requires a strict opt-in Dutch sites tend to use splash pages or white boxes before letting you see their content – such as this example from RTL Netherlands. RTL is a pan European company, their other sites do not force this on their visitors – but the cookie law implementation varies across the EU.

The cookies we set on our site serve three functions;

  1. remembers which language you want to use to read the site
  2. remembers your response to a disclaimer (we include some information that is not supposed to be for the US market)
  3. collect (anonymous) data on your visit so we can improve the site

We didn’t want to force people to opt-in so it’s optional. Not surprisingly most people don’t which means that on return visits they may need to re select their language, and they may need answer the disclaimer multiple times – and this applies to visitors outside the EU. We now do not get enough data to analyse the site.

There’s no good solution to this; either we annoy visitors with the forced opt-in, or we don’t collect enough data to analyse our site, or we don’t comply with the law. It’s a frustrating situation to be in. Particularly as we know from other research that 90% of visitors will leave the cookie acceptance on the default setting – even if that is the highest setting.

The ICO, the organisation responsible for the enforcement of the cookie law in the UK, announced a change to their use of cookies earlier this year, effectively moving to an opt-out model. For UK sites the cookie law is a vestigial form; you need to disclose how you use cookies but specific opt in is not required. Here’s a helpful timeline of the developments in the UK.

For Dutch companies the requirement remains unchanged, full opt-in is required and companies must collect proof that visitors have opted-in (that’s the extra data we’re now collecting about you). There’s no indication from OPTA (the Independent Post and Telecommunications Authority – the organisation charged with enforcing the law – part of the Authority for Consumers and Markets), that any change is planned. However Dutch site “Marketingfacts” reported that a bill amending the cookie law was presented to parliament on 20 May (Article only in Dutch). The proposed changes would allow analytical cookies and those needed for the operation of the site to be set based on implied consent provided the data collected did not have an impact on privacy.

The bill has been through a consultancy phase and it will now be up to the minister to decide whether to submit the bill to the lower house. Like many of those working in digital industries I am hoping this bill goes through.