February’s Mobile Dorset saw one of our biggest numbers of attendees yet - all attracted to the excellent talk by Manuel on eye tracking, and of course the free pizza and chat with like minded mobile professionals! Here’s what happened …

Talk - Eye Tracking In Mobile Apps

We were delighted to hear from Manuel Martin Salvador, part of the R&D team at Base, who gave a talk on Eye Tracking in Mobile Apps. Manuel explained how advances in machine learning and the compute power of mobile devices meant it was now feasible to track what part of the screen a user might be looking at, rather than require specialist equipment costing hundreds of pounds. They tested this by building an app to train a model by having the user follow a ball around a screen, resulting in an accuracy that was much better at detecting what part of the screen the user was looking at than a human could!

Manuel shared with us the results, and continued to discuss what practical uses this might have, particularly in the area of accessibility and of course, advertising.

For more details, check out Manuel’s blog post on the subject, or pull the Github project for yourself and help improve it!


After the usual voting on topics again we spent the rest of the time discussing the following items:

Mitigating the risk of relying on 3rd party services

  • Following the shutdown of Parse, we discussed peoples thoughts on relying on services like this.
  • In particular a distinction was made between paid for services and those offered for free, or at least where the company relies on the income to survive - such companies seem less likely to disappear.
  • From a technical perspective, ‘wrapping’ the service in an interface so that it can be swapped out later for a different service later on was a sensible strategy.
  • However, need to take care not to blow your budget on that activity at the expense of other work.
  • Ultimately decisions should not be made lightly when choosing a service, and the Developer has a responsibility to raise any service they are using to allow contingencies to be discussed.

Imposter Syndrome

  • One of our members was asking how to spot and succeed when faced with Imposter Syndrome - where a high-achieving individual cannot recognise their accomplishments and fears being exposed as a ‘fraud’ and not deserving of any success.
  • Moreover, does this trait contribute to being ‘looked over’ for promotion or other rewards and what can be done about it?
  • One suggestion was to keep a work log of successes to be able to refer to later on when needed.
  • Also while tough, trying to highlight your own successes with leadership and the team would ensure you weren’t ‘forgotten’, though how would this work in flat hierarchical structures or if you can’t recognise them to start with.
  • Hence maybe it’s not for the individual to work out alone, the leadership should be looking for this and ensuring the right individuals are rewarded accordingly.

How to make a mark on the app stores

  • You’ve published an App to the stores - what is the right way to promote it?
  • Getting in the charts seem to give short term success, and some teams have used bot nets to achieve this.
  • Also, press attention can create a spike in downloads, or ‘free app for a day’ promotions.
  • But anyone who has had this has seen the success be very short lived, and certainly not enough to make a living from.
  • These days people consider the app store just a distribution method, and so marketing attention should be put into more direct activities to earn users.
  • Someone asked for any experience with the two classic types of revenue models - should you charge up front or use in app purchasing?
  • As usual, the answer depends - what are the target users used to? What type of service is on offer?
  • Some people have moved on to revenue models based around affiliate fees - passing users on to somewhere else and taking a cut.

Prevent Stealing Ideas

  • Finally we discussed how to prevent an idea being stolen while you look for a Developer to partner with.
  • NDA’s are the common method for this, but might be too heavy handed early on, and some people refuse to sign without knowing a little bit first, or actually being under contract.
  • But we were reminded of the words of Eric Reis in Lean Startup whereby he posits that no competitor would have the time or knowledge required to build the idea before you would yourself, so to focus on getting to market (using an MVP, of course).
  • Indeed, once launched anyone can compete so maybe better to focus on launching first and staying ahead than any defensive move.

Topics we didn’t get to

  • Ian Livingstone starting a school in Bournemouth
  • What’s next for mobile
  • Best practice for colour blindness
  • Printed cards for prototyping
  • State of platform market - Windows dead?
  • How to dynamically change strings.xml
  • How to start learning Rx Java


Another great chat and having a speaker was well received so we’ll be looking to have another talk in March and are just working out the details, but grab a ticket now and follow us on Twitter to get more details as they occur!