Scribd – the wrong way to use Facebook personalisation

Scribd – Where the World Comes to Read

Scribd is the world’s largest social reading and publishing company. We’ve made it easy to share and discover entertaining, informative and original written content across the web and mobile devices. Our vision is to liberate the written word, to connect people with the information and ideas that matter most to them.

Or, alternatively:

Scribd – Where the World Comes to Look at Your Facebook Data.

Just a few weeks ago I wrote a post about how data was the future – and the past – of the internet, and how it could be used to really improve the user experience. Sods law states that just days later I’m faced with an example of how to use it to destroy the user experience.

I read articles every day, everything from white papers to blogs to news stories. It was only a matter of time until the document I wanted to read was hosted on Scribd. I don’t have a problem with that, and initially it all seemed pretty good. Yes, I needed to log in to get hold of the document, but Facebook Connect was available and that should have eased the process. All cool so far.

It was here that things seemed to take a stroll downhill.

First, there was the odd policy of having to upload a document before I could download a document from the site. At this point I didn’t have anything ready to share, so I declined and decided to read the document online instead. A bit like SlideShare. Odd, usually you would give users an opportunity to try a service before deciding to participate in it, but not a problem.

Scribd – I get a few emails

Then came the emails. 9 emails. All telling me that I was being followed by someone. Impressive I thought… for approximately two seconds. My next thought was ‘unlikely’. Unlikely that two of my Facebook friends happened to be online at the same time and had both seen me join the service. My spider-senses were tingling.

An hour later they were deafening. Especially as my wife was following me on Scribd. She is on Facebook, she also writes a blog, but she most definitely wasn’t on Scribd. Being inquisitive, I clicked to view her profile. Apparently she joined in January 2011 – before I did.

Scribd – Julie’s ‘profile’ – note the joining date

So not only is Scribd grabbing my friend’s information from my profile, its also creating accounts ready for them. It’s just plain wrong. That’s their data, not mine, and they haven’t given permission for Scribd to hold their data; I’ve given Scribd permission to hold and use mine. And yes, I do see their name and picture being used in conjunction with an account as being a breach of that trust. They are associating their service with someone who they don’t know and have not had any interactions with.

And at no point have I had the option to opt-out of this happening.

Being kind, very kind, I can see a reason why they might act this way. After all, if you’re a Facebook Partner for the Personalisation product, you want things to look good. But if your service only has a small user base then the chance of a group of people you know stumbling across your account  – or even more unlikely, a group of people you don’t know stumbling across your account – and choosing to follow you, is very small. I can imagine the product brainstorming meeting that morning: “Hey, I’ve got a great idea, let’s just create a load of accounts every time someone logs in using Facebook! Brilliant! That’s thinking out of the box!”. Unfortunately, yes, it is out of the box, and for all the wrong reasons.

I’m not the only one who has had a similar reaction to this behaviour. Rohit Mishra made similar points in his blog post in February; although he found out what was happening in a different way. They’ve also got called out in Wired last September. You would have thought that they had learned their lesson by now, but obviously not.

Data sharing – executed correctly – has the ability to create immersive and rewarding social experiences. I think we should all take a look at Scribd so we know exactly how not to do it.

**** UPDATE 29th September 2011 ****

It was brought to my attention by a tweet from Laurence Buchanan (below) that Scribd isn’t the only site in Facebook’s Instant Personalisation program. Rotten Tomatoes also takes the same approach. You can find out more at Techcrunch about the initial partnership.

To check the veracity of the claims, I visited the Rotten Tomatoes site. Without logging in or creating an account (this will become important in a moment) I did see my friends movie likes and dislikes. Imagine my surprise when, upon visiting the application settings page in Facebook, I was greeted with the following.

Rotten Tomatoes application setting in Facebook

I really don’t remember giving them permission to do anything…

If you want to read more about this, I’ve looked further into Facebook Instant Personalisation here.



Why digital marketing is like my local

Okay, I’ve got a confession to make. A few years back I used to be the captain of a darts team, playing every Tuesday in a league. I wasn’t very good, but as I was capable of organising things I got the job by default.

Putting that to one side, being captain of the team and a regular at the pub had its advantages. I was well-known – in fact I still am, even though I haven’t been a regular for years – and that brought its privileges. They knew what I drank, I’d get my name up on the board for a game as soon as I walked in, and I was never short of someone to chat to.

So being a regular at a pub is like digital marketing.

Sign for The Slaughtered Lamb pub in 'American Werewolf in London'
Not my local pub... there's no werewolves there as far as I know...

Are you still with me? I know that last statement seems a little bit of a stretch. I promise I’ll make up for it.

Every visit I made, the people there learned a little bit more about me. They collected data, building up a picture of my likes and dislikes, my past and my present. It’s that data that they used to create a welcoming, but not too intrusive experience every time I went there (whether they knew it or not).

A few weeks back at SXSW, Reid Hoffman, the founder of business social networking site LinkedIn, said that Web 3.0 was all about the data.  But the thing is, the web – and digital marketing – has always been about the data. From its humble beginnings at CERN to the present day the web has been about the exchange of information; only the type of data has changed, from documents to social and personal. As users, we’re now involved in a quid pro quo arrangement that says ‘you give me some of your data and I’ll share my personal data with you’.

It’s true to say that offline marketing is about the data as well; the very existence of CRM systems pays testimony to that fact. But it’s the ease by which we can collect data online that is so compelling. There’s no manual collection and processing of returned direct mail, no manual data entry every time a phone call is completed. It’s so easy to collect data in a digital environment that we’re really getting spoilt. Despite concerns over privacy – which Facebook is never far away from – Internet users are more and more ready to share information with the established social networks. Using tools such as Facebook’s Graph API we can start to use this data too, in our own little quid pro quo with the social networks. Facebook understands a little more about the interests of its user base and we get a little bit of demographic information back in return.

The important question now is ‘How do we collect and use data in the right way?’

Let’s go back to my local pub and the experience that was created. It was welcoming, reassuring, but not obtrusive. It was a relationship based on familiarity, not stalking. Digital marketing can create the same relationship if it’s approached correctly.

Digital Marketing Data Do’s

  1. Make collecting data easy – your user knows that you’re going to want to get information from them at some point, they’re not daft. When you do, make sure it is as easy as possible. Give them the opportunity to use existing logins and then supplement that if you have to. Open the door and welcome your users in.
  2. Give something back – Digital marketing is about give and take. If you want users to return to your site, and by doing so give you the opportunity to learn more about them, you need to give them a reason to do so. Invest in good quality content for your core web presence. Great design is the icing on the cake, but users don’t come back simply to look at how you’ve styled your navigation elements, content is the key. If you can’t create truly original content then at least curate and comment to add value to existing content.
  3. Personalise – don’t collect data if you’re not going to do anything with it. Use the information you collect to personalise the web experience. We’re not talking targeted ads here, just the ability to present relevant and engaging information to your audience. For example, use location to filter or alter content so that it is meaningful: if you present a list of shop addresses, place them in order of distance from the user’s location or alternatively just move the nearest shop to a featured position at the top of the list.
  4. Respect privacy – not everyone will want to get to know you. Ensure that you make it easy for people to disengage. They want to remove their details from your database? Do it straight away and let them know, preferably whilst they’re online and with you. Transparency creates trust, and you never know, they may come back.

Digital Marketing Data Don’ts

  1. Don’t be obtrusive – “Good Internet companies do not ambush their users,” said Reid Hoffman in the interview at SXSW. He’s absolutely right. Personalisation is a great tool for engaging and keeping users, but played too hard it becomes just a little creepy. Be subtle about collecting information.
  2. Don’t lose the data! – the last two weeks have been full of stories about lost data, the most prominent of which has been Sony’s PSN debacle. The incident is likely to cost the company around $24 billion in compensation and lost revenue, but it will also have a massive impact on their userbase’s relationship with the brand. How much additional revenue will be lost due to people’s reticence to spend money online and trust Sony with credit card data? Only PSN’s position as the only way to play multi-player games on the PS3 will hold it steady. If it had been more dispensible the impact may have ended up being much more severe; losing or exposing your user’s data is unforgiveable.
Digital marketing presents us with an amazing opportunity to build lasting relationships with consumers online, and the potential rewards are huge, but the onus is on marketers and their clients to set the benchmarks. We have to show respect for our users and engage on their terms. Only then can we realise the benefits of true relationship marketing and keep our consumers coming back again and again.

Let’s be Open about things

There is a misconception that to get the most out of Facebook, you have to be on Facebook.

Not true.

Certainly you can create a fan page or a company page, and get your followers to ‘like’ you, but beyond the numbers what are you getting? Social connections don’t automatically mean that you understand your audience better.

In the past, we relied on producing great content to attract our users, and with the promise of more exclusive content, tried to convince them to fill in cumbersome login forms so that we could understand our target market better. It’s not a great tactic and overall it doesn’t produce great results.

There is a better way.

OpenID LogoOver the last few years there has been a move to try and simplify the experience of using the web, and in particular how we store and supply our information to different sites and services. At the head of this move is a technology known as OpenID, governed by the OpenID foundation, a group that consists of representatives from Microsoft, Facebook, Google, Yahoo and IBM.

The premise of OpenID is that web users can create one login and then use it anywhere on the web. In fact, most of us have an OpenID without even knowing it. Google and Yahoo accounts are both OpenIDs; although Facebook and Twitter aren’t. But even so, services such as Facebook Connect and Twitter’s implementation of XAuth provide a similar experience to end users.

Initially, there were more Chiefs than Indians, with all the big web companies wanting to sign up to the programme as OpenID providers and not enough sites accepting them as logins. In hindsight this is unsurprising, as control of information is the very essence of the big web players such as Google and Facebook. But over time this has changed. Now there are many sites where these IDs are accepted, from Fox News, Digg and FriendFeed to Kmart and Sears, and the good news is that it’s simple to get your site accepting them too.

So how does this benefit our users? When they arrive at our site they can choose to login with a pre-existing account, so there’s no need to enter a username, choose a password and go through the rigmarole of activating your account. They simply enter their email address and password and they’re done. There’s a great explanation of how this works at

Login to Ficly

In return for making things simpler for users, we get back some additional profile data from them when they log in. The data you receive changes by provider and dependent on consent from the user, but can contain data such as their name, gender, birthday, email, time zone, address, postcode, country, web page, contacts and phone number. All this without filling in a registration form.

Of course, there’s nothing to stop you from adding an additional couple of questions to the login process in order to capture more client-specific data as and when necessary.

So what does this mean for marketers?

Utilising services such as OpenID and Facebook Connect we’re lowering the barriers to engagement for our targeted users. This, along with positive brand association with services they already trust with their data, will make them feel more comfortable about sharing their information with us, increasing registration rates and the amount of data captured. And the more data we capture, the more effectively we can target.

It’s in our interests to start being Open.