Thursday, January 30, 2014

Nobody will read this anyway

In my first year at university I studied journalism. One of the few lectures I care to remember was on writing for newspapers. The lecturer told us to throw out everything we knew about writing, nobody wanted to read it. I liked the sound of that. We needed to tell the entire story in the first 25 words. Interesting. It definitely works, but right now I'm ignoring this rule.

I’ve been working on internet projects for 15 years, and over that time I’m built a fairly lengthy list of lessons learnt. From this experience, I’ve identified three key areas which consistently create a greater than planned share of effort. I’m not going to give away all of these lessons in one article, but I will give you my number three; Content.

Content is always seen as the easy piece of a technology project, but more often than not I see projects struggling to deliver appropriate content to schedule. Why? Most digital projects do not adequately consider the new content features and content areas that they have now created. It's a bit like moving your old furniture into a new house... it doesn't always fit.
I’ve been lucky enough to be involved in our industry from the dot com days and witnessed the first digital forays by many companies and brands. To me, there have been three key generations of digital content creation:

1.       Content for web – at the very start of the internet, companies were simply taking content from their print brochures and collateral and moving it online. As this generation matured, we identified the different consumption patterns of digital users and the need to tailor content specifically for web.

2.       User generated content – this was a complete disruption to the existing content model where information was pushed from organisations to end users. The internet gave users the ability to quickly and easy provide commentary, imagery and video. This was the dawn of social media

3.       Content for mobile devices – with now more mobile phones than toothbrushes in the world, the current challenge for content owners are is how to tailor content for small mobile screens, tablets, televisions and the continued growth of user generated content. Welcome to the beginning of the responsive web.

I believe we are now entering the next generation of content, artificial intelligence. Content needs to be dynamic, personalised based on a consumer’s preference, devices, behaviour, physical location, time of day, attitude and myriad of other personalisation possibilities. There are far too many permutations for any content author to effectively deliver this mass personalised experience. Perhaps you would like to provide a special offer to all returning customers located in Australia, accessing your site at 5pm on a Wednesday after reading your email newsletter and who previously made a purchase from your  site?. This sort of complexity has historically been different for non-technical people to achieve, but thanks to a growing breed of artificial intelligence inspired customer experience platforms, it’s actually very easy to target content to the most specific of audiences.
We are handing control to software tools and platforms that learn from past visits, and optimise the site to be more effective. Is it any wonder Google is buying companies such as DeepMind. They get it.

So to appease my university lecturer here’s my “25 words or less” summary I should have put at the start of this blog.
“One of the top project risks to technology projects is content. Not simply writing, but considering how consumers context changes the information they really want.”

Wednesday, October 30, 2013

Who let the dogs out?

It was the year 2000, perhaps this was the real millennium bug, but anyone who had a radio will know the irritatingly catchy song from the Baha Men - who let the dogs out? In writing this article I read a few too many blogs and forums with people trying to establish the true meaning of this ridiculous song. I'm not going to weigh in on that discussion, but the title and chorus is a catch cry I'll unfortunately never forget. 

One of the questions I've been asked many times by clients is how to best measure success of digital campaigns and activities. In a world of digital and data, you'd think establishing metrics is easy… but it's not. It may be easy to create metrics, but are they the right ones?

Most management consultants love a 2 x 2 matrix. If you don’t know what one is, you are in for a treat. Think of a very simple two axis chart that measures cost or investment on one axis, with the other axis measuring value or return. Or here’s one I prepared earlier.


Now on this matrix, I want to plot all your digital campaigns or activities based on their relative investment cost and value returned. You will most likely find it easy to identify cost, but if you are like most organisations you will find value a little harder to define. Is value based on visits, clicks, downloads, subscriptions, registrations, sales, the list goes on? This is a critical decision, but for the purpose of our matrix, just pick what makes sense to you. If you are unsure of the value generated, it’s likely the campaign is in the lower end of the value axis.

So what does yours chart look like? My guess, most of your campaigns will be in quadrants 1 or 3. Imagine my surprise when most companies I talk to end up with most campaigns in the top left quadrants?! Simply put, this translates to high cost, low returns. By any measure, a poor result. I once heard a senior leader from the software company Sitecore describes these campaigns as "dogs". Why do we continue to operate costly campaigns with poor returns? Or worse still, costly campaigns with no clear idea of their return? The matrix below is a simple labelling of the types of campaigns; sleepers – for the campaigns with low cost and low returns, heroes – the campaigns with low cost and high returns, dogs – those with high cost and lower returns, and whales – take a lot of resources, but do deliver high returns.


On reflection, the main issue is that it is very difficult to define or agree on the true return or value generated. 

Understandably, every site is as different as the businesses they represent. Measuring value seems simple with ecommerce orientated websites, but far more complicated for a site where the purpose is generating sales leads, or delivering content. 

I can't take credit for the following model, but these three steps make a lot of sense to me:

  1. Establish a unique and relevant value scale for your website - every website is different, and every activity and interaction on your website has a different benefit to your organisation. For an automotive manufacturer, requesting a brochure is not as important as booking a test drive. It doesn't matter what the value scale is, it just needs to appropriately weight the most important interactions for your business.
  2. Each campaign needs to identify the key target outcomes (and their associated value) - every campaign should have a clear call to action and also easily attribute website traffic as its origin.
  3. Measure each campaign - based on a very simple formula, we now multiply the number of goals or interactions triggered by the value of each interaction. For example, suppose an automotive company ran a campaign that generated 10 brochure requests (relative value scale of 20), we would say 10 x 20 = 200 value points. If a second campaign delivered 15 test drives (with a relative value scale of 100), that campaign achieved 15 x 100 = 1,500 value points. These may initially seem like random numbers, but it's a score of relativity. With a common and agreed value scale, we can now very easily compare campaigns and more importantly identify those campaigns providing fantastic value, and stop those dogs.
So next time you hear "who let the dogs out", I hope you can answer at least identify the value of your campaigns, and confidently say that you are investing in heroes. 

Sunday, October 6, 2013

All clicks are equal, but some are more equal than others


Many years ago I was presenting the monthly results of an email campaign to the Marketing Director of a global manufacturing company. I was very proud of my colourful campaign report and results, so I proceeded to share graphs and “insights” on all those important benchmarks synonymous with email marketing – open rates, bounces, clicks and unsubscribes. Now this particular client had always been a reluctant participant in any form of digital marketing and hurriedly brought the meeting to a close. At the end of the meeting he thanked me for me time and asked a simple question, a question which changed my views on digital marketing forever: “Out of interest,” he said, “who were the three guys who unsubscribed from your email campaign?”

What on earth did he mean? Who were these three guys? Who cares?! Clearly they were three customers with poor taste, no longer interested in my beautifully designed email or well written content. Who cares who they are?!

I smiled, left the meeting, hoping to ignore his question and bemoaned my bad luck of working with clients who clearly didn’t “get” digital. By the time I got back to my office, I was desperate to find the answer. I didn’t take long to dig through the email reports, find the offending email addresses and track them back to the customer database. Done. But that wasn’t the end of it. Fortunately, or unfortunately, I also had access to this company’s past three years of sales data. So I linked this information together and created a quick profile on my unsubscribers. Thankfully two of these customers had never bought from my client, and this was actually the first email they received. Who cares if they left, clearly no loss. But the third guy, different story. This third customer… who was this guy? Turns out he was more than some random customer, he was a lifelong customer who had purchased literally tens of thousands of dollars of products. If there was an award for most valuable customer, this guy would top the list. What is going on?!!!!!

Shocked and rocked with this new perspective, I re-examined my now questionable campaign report. Forty percent open rate, 20% click throughs, and 3 unsubscribes. Great stats, but who were these people? What if I asked “who” for every click, bounce and open? I had been producing seemingly wonderful (substitute: meaningless) reports without the most important dimension; who. This is the digital equivalent of counting revenue based on the number of monetary notes received and ignoring the actual value on each note.

The shocking thing is, we still do it. With the explosion in digital devices and big data, we have stats for everything; Facebook likes, followers, tweets, app downloads, page visits and of course email clicks. By aggregating these as some sort of score or badges of success, we are treating all these interactions as equal, ignoring their independent value and most importantly turning our back on who generated them.

Every week I see numerous digital marketing reports. All of them are interesting, but many of them miss the most important dimension. Whenever I’m asked to comment on one of these reports, I always ask one question in return: “can you tell me who these customers are?”

 

Monday, September 23, 2013

"Ask me twice"... starting a new blog


I’ve tried many times to write a blog. This may simply be the next failed attempt. It wasn’t time or technology which I found the barrier, rather the challenge of finding topics to write. I enjoy writing, I’ve found it difficult to know what to write.

Some time back I asked a colleague who I think to be a rather successful blogger how he kept creating such a constant and interesting stream of posts (“interesting” being a completely subjective concept). His answer and approach is simple. Whenever he is asked the same question by two or more colleagues or clients, he posts his answer. He figured that if two people had asked, more would be interested.

So I plan to road test his idea. Ask me twice, and I’ll blog my answer. My disclaimer of course is that anything I blog about will be heavily filtered through the lens of being a digital technology geek. So when my kids ask me “why” twenty times a day, I won’t blog about it.

Over the past few months I’ve been building a list of questions or topics which seem to be reoccurring. Some of them relate to my work, all of them relate to digital.

Hope you enjoy.