Return to site

Mobile Fix – October 14

IVR and ChatBots

Many years ago, when the information superhighway was the future  - and the web was an obscure academic project – there was a now almost forgotten tech network delivering content and services. Sometime called Audiotext accessing content via premium rate phone lines was highly successful and made some people a lot of money.

Many of the early players on the original Walled Gardens– Prodigy, AOl and Compuserve – were people who had learned what worked in audiotext. And interim technologies like Minitel in France and Teletext and Ceefax in the UK were influenced in a similar way.

One persistent reminder of these days is Interactive Voice Response – the interface used to navigate the content menu. Still prevalent in call centres around the world IVR was best summarised by The Simpsons

And quite apparent in the way many Bots work today.

Best practice is enabling a user to get the answer they need, rather than just moving them through a restricted set of options. This came to mind at a fascinating workshop organised by Amazon to help people, understand how they can use Alexa - and in particular  develop Skills. Skills are the ‘apps’ of Alexa and already a number of UK brands have launched – from a Daily Mail news update to EON bill checking. Providing real utility and a great user experience is the key here and the Amazon team emphasised that the planning up front is probably more important than the development back end.

Essentially Alexa, Siri and Google Allo are new ways to initiate a search, but if your response is little more than voicing the top results, and the equivalent of please hold for an operator, you have missed a trick. Our theory is that chat bots will quickly morph from text input into voice input – remember over 20% of mobile searches are already voice – so these opportunities re going to collide. We are in the early days of voice and chat. ( Benedict Evans has written a good blog post on this topic)

Amazon Prime

The support for Alexa and their other Hardware devices shows how focused Amazon is on colonizing the home – but we wouldn’t be surprised to see Amazon try again with a smartphone. Their last one wasn’t that bad but turned into something of a debacle. But with a Samsung Note 7 shaped gap in the market, could they try again and make a success of a voice focused device?

Given that their objective is signing up people to Prime we suspect they could end up giving the Alexa away to their customers as an incentive to join. Could similar economics apply and lead to a heavily subsidised Amazon smartphone? Their partnership with Blu hints at this sort of direction

Estimates suggest there are around 63m Prime customers so this data showing 5% of them spend over $10k a year means that’s 3 million people? Really?

Other Amazon news is that they are opening more retail stores and introducing pick up points – focused on their growing US grocery business. There is little data on how they are doing on the UK with Fresh – in our experience they seem to have the same issues as the competitors in terms of deliveries and missing items.

One of our favourite books is still Net Gain by ex Mckinsey guru John Hagel. In the book from the late 90s he talks about building a community and then finding products for that community. Amazon is demonstrating this approach works really well.


We had to miss the always excellent IAB Engage this week but we heard great reports, Not least of Fix friend Claire Valoti making the first public presentation by Snapchat in Europe. Stating that they think of themselves as a Camera company, the stats shared included the fact they have over 10m daily users in the UK – 77% of which are over 18.

There is always lot of news around Snap. They are dipping their toes into programmatic and looking at Search.

The Mobile Wrench

In her talk Claire Valoti made the excellent point that we need to work out what mobile ads should look like.

We agree but we also need to think through what mobile content looks like. Whilst the majority of Snapchat users create content every day, most mobile content is repurposed from elsewhere – and /or isn’t that different from the content on desktop.

We keep coming back to DW Griffiths and his Wrench.

In his 1911 film the Lonedale Operator, Griffiths introduced the close up – showing what had been thought to be a gun was actually a wrench – and defined Movies as a new medium. Until that point Movies were essentially theatre, captured on film.

As Snapchat look to Hollywood for new content Facebook have an interesting new format that could help define the medium. As a promotion for a new TV show Facebook users will be given clues and asked to help solve a mystery.

Sounds ambitious and supports our view that episodic content is ideally suited to Facebook and other social channels - and we are surprised there isn’t more experimentation. The Soap Opera came about as brands invested in new content forms to drive new media audiences. A similar approach for mobile would be a better use of budget than importing a 30 second TV commercial.

Quick Reads

Facebook launch Workplace – the first of their Products to be developed by the London team - and give Slack and Box etc some stiff competition.

As potential buyers for Twitter make their excuses and walk away the current leadership seek to rally the team

Tracking store visits driven by mobile ads is a major focus for Google Facebook et al and Fix friends at On Device Research are looking at the area too.

A very good piece on how markets can be segmented – Cutters, Shavers and Nevers from GroupMs Rob Norman

Apple make it easier to opt out of ad targeting And more people seem to be taking this option since iOS10 was launched. But their isn’t much upside  - you see the same number of ads but they are less relevant

Finally Fix Friend Faris weighs in on personalized ads and makes the smart point that Personalized ads are an oxymoron. We agree – the promise of dynamic creative can be overstated and no one wants Minority Report ads.

But we do believe that Ads can be improved with data.

Just like the computer science saying that computers should never ask something it can work out the answer to, ads should never tell you about something you won’t be interested in.

If a Hotel app knows you are in Leeds it should tell you about Leeds hotels. If an airline knows from your frequent flyer programme you tend to fly Business it should never show you an Economy fare.

Adtech lets us make smart judgements about the people seeing our ads. If we can recognize someone, our ads should never treat them as a stranger. But we never need to be too presumptuous. We should use the data to give the most appropriate message. A former colleague talked of Advertising so good it’s a Service.

Improving ads with data is good for the user and the advertiser.

All Posts

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!