Using Data Analytics To Improve App Development

3 areas for data analytics and optimisation

There are three key areas where we use data analytics to inform what we do:

Conception – How data helps inform the initial idea

Design and Build – Using analytics throughout production

Optimisation – Ongoing measurement and iterative improvements

Conception

Data analytics are important right from the off. 

During initial discovery phases for example, a lot of critical decision-making has to take place to work out what our clients’ mobile roadmaps should look like – the qualitative and quantitative research we do at this stage ensures the validity of those choices. 

One of the first questions that we always ask our partners when beginning on a project is “why do you want an app?”. 

Generally speaking, the most common reason is that businesses want to create an experience that just isn’t possible with web – or so they think.

We’re strong believers in the idea that tech for the sake of it just doesn’t work – which is why if a client comes to us wanting an app, we scale the conversation right back to “what’s the problem you’re trying to solve?” to see if that type of solution is suitable.

Where it is, mobile apps open up an entirely new direct marketing channel, where you can feed information to your users. You can build a brand, improve customer engagement and ultimately stand out from your competition.

But only if you start off in the right way – some key tips:

1. Research the potential: Running competitor research is a must, allowing you to determine whether you can penetrate the app marketplace with your new proposition. Identify and examine your competitors’ online presence to help determine how you can differentiate. 

2. User research: This is integral to determining whether your app is going to fulfil a demand – you need to be certain it will provide value. We always interview current and prospective end-users to find out whether the solution we had in mind is the most suitable, and if so, let that feedback inform the functionality.

3. Leverage existing data: If you want to build a mobile app but already have a website, you should already have a wealth of web data analytics that can help inform whether your proposed solution is the most suitable. Think about what data you already have at your disposal and how you can turn it into a valuable resource to inform future digital products along your roadmap.

4. Create more: Demographic, audience and behavioural data analytics can all provide a great starting point for confirming the viability of your new proposition. If you don’t have any, why not run a quick poll or survey to find out more?

Design and Build

Regardless of what we’re building, we always try to utilise data wherever we can to help inform our decisions. 

It might be easy to get lost in reams of it, but providing you utilise any available data for the right reasons, in the right way, the chances are far higher of the end result hitting your objectives.

So, let’s start with design. 

First, you should always use data to inform and not lead the design process, in order for your new product to bloom. 

“If I had asked people what they wanted, they would have said faster horses” – the classic Henry Ford quote. Data can be invaluable to your design team, but it’s essentially the driving instructor, not the driver. 

That being said, data can help identify what features and functionality make up your new app. For example, let’s say you’re running an online clothing store and want to develop an app. You know from your current website that most of your users utilise the ‘sizing guide’ feature – and doing so significantly increases the chance of a purchase. 

You know this feature is a showstopper, so when building your app, it’s a must-have element, and plenty of thought needs to be put into how it’s designed. 

You could also then use qualitative data – like running usability testing on the current product page – in order to help further improve the sizing feature functionality on the app. 

Next, the build. 

The development cycle is influenced first and foremost by tracking analytics. 

This usually starts with a specification detailing everything that both Apadmi and our partners wish to be setup once launched. 

A specification acts as an agreement between the analysts and our partners, whilst also providing the developers with a set schema on what’s been agreed. 

It’s critical that our engineers are part of this stage – they’ll almost always have an idea on how to make it more efficient and future-proof.

Optimisation

So, you’ve got your shiny new solution with a full analytics setup and it’s performing as expected. 

Well – there’s always room for improvement, and that’s where data-informed optimisation comes in, and there are a few steps we take to do this…

1. 1. Simplify your ongoing analysis – we’ve developed our own solution (called Operational Success) which provides our clients with a dashboard that actively measures KPIs and ensures that all solutions are running as expected. It also means we can quickly generate reports from that ongoing data set via an online portal around the performance and success of that solution.

   2. Follow the trends – once we have that baseline of performance, it makes life easier when we want to spot any trends and follow up with deep dives into our analytics. At Apadmi, we pride ourselves on being able to utilise data to inform larger business-wide decisions. Check out how we helped specialist nurses save 766 hours through improved productivity using data!

   3. Collaborate – when discovering opportunities in our datasets, it’s time to get the team together. Product owners, developers, designers, partners – whoever we need in the room to make things better. A core part of our business is that when the solution doesn’t exist to fix or enhance something, we create it, bringing a more technical approach to our sessions.

   4. Validate – we won’t simply roll out solutions though – we need to validate them first. Some larger scale changes may undergo moderated usability testing, where we can better gauge the reception to any new functionality. Alternatively, we may also utilise unmoderated remote user testing to gain quicker and more iterative feedback. A split, A/B/n or multivariate test, upgrade options – depending on the scale and complexity of the build, there are multiple testing tactics we might use to determine genuine user reactions to changes. Regardless, we’ll continually utilise data in order to determine the success of whatever changes we make, and therefore inform future decisions on that functionality. 

 We’re experts in the mobile ecosystem – which is why we also need to be experts when it comes to data.

More specifically, knowing when and how to leverage every data resource at our disposal to build great mobile solutions – and continually stop, analyse, test and improve so that their performance keeps up to date with expectations of the marketplace.