Skip to main content

Your web browser is out of date. Update your browser for more security, speed and the best experience on this site.

Update your browser

We use cookies to make interactions with our websites and services easy and meaningful, to better understand how they are used and to tailor advertising. By using our website or clicking “ALLOW”, you are agreeing to our Privacy Policy.


It's a big data world

There's one building block at the core of every startup, whether you're planning the next great consumer tech company or a powerhouse B2B SaaS startup — data. For this part of our LiftOff series, we sat down with our in-house data expert, Accelerator Centre mentor Steven Fyke.

December 14, 2021

There's one building block at the core of every startup, whether you're planning the next great consumer tech company or a powerhouse B2B SaaS startup — data. From marketing to product development to the product itself, collecting, analyzing, and putting data to work is a massive undertaking.

For this part of our LiftOff series, we sat down with our in-house data expert, Accelerator Centre mentor Steven Fyke. In addition to being a mentor, Steven is the founder and President of SnapPea Design, a Kitchener-based design shop that helps startups and large corporations with product design challenges. He's also the founder of Furna, a consumer tech startup that has created a vaporizer that uses pre-packable, swappable concentrate and dry herb ovens. 

Steven wants you to know that while we live in a big data world, it's critical to understand what to collect, what to keep, and how to represent data in a compelling way. 

Many data points can be collected — and collecting as much data as possible is valuable. But it's more than about what you collect. It's about how you process it. There is no easy straight forward answer most of the time. Like most things, Steven believes in trying, testing and refining as the product and needs evolve.

You have to understand what the impact is from a data collection standpoint. Are you putting it into a data lake and post-processing it or are you using it as a real-time analytic. It gets really convoluted very quickly," Steven said.

What data do you collect?

We asked Steven to give a non-tech example using a neighbourhood auto repair shop to understand these questions better. We all know an ounce of prevention is worth a pound of cure, so what data would a repair shop need to collect to provide a predictive maintenance service for their customers?

Steven said first, you need to understand the need and value to your business and customers. If your goal is to increase repeat visits from customers, you could start with digging into the data to see when you should start calling customers to book appointments.

‍It doesn’t have to start with massive infrastructure and IoT devices everywhere. It could just start with data on their driving frequency. You don't have a direct connection into their car (yet), so you could start recording data every time they show up such as how many kilometers they've traveled. That data will start to give you a model that can predict when certain maintenance procedures should be done, based on routine maintenance" Steven said.

How do you use the data?

The type of relationship you have with your customer can also affect the types of data you can collect. Continuing with the auto repair shop example, Steven moved from passively collecting data when customers come into real-time data from vehicle sensors. All automobiles have an OBD port that connects to a data terminal in the repair shop — and that data can be used in conjunction with other data sources to provide even deeper insights.

Let's say one of your customers drives a pickup truck. You could track how many kilometers it's been driven, where it was driven, the weather, and all kinds of other seemingly unrelated things to build unique insights when compared to similar vehicles,” Steven said. “You have all this data from a specific truck that enables you to do comparisons between different trucks and give you predictions on what work will be needed. If you have limitless resources, you can create tons of interesting analytics but having a hypothesis on the value of each piece of data helps to narrow your focus.

What data do you keep?

‍Collecting data has costs — in the methods you use to collect it, the tools you use to process it, and the storage costs.

Doing data collection on the fly gets expensive. Now you have to store all that data. You need to ask yourself how valuable it is to store every detail," Steven said.

For the vehicle data, one way would be to create analytics metrics. You could create a score based on the RPMs, torque, and CO2 output. But once you condense those data points into a singular one, what raw data do you need to keep?

What business needs do you need to keep the raw data? What parts? How do you want to process it? There's a lot of questions and unknowns in there because data storage costs money," Steven said.

Putting data to use

Data can also be a powerful differentiator for your product — even if your product hits the market before you have data. Let's say you're building a running app — a market that is already highly saturated and competitive. How do you differentiate yourself to attract, retain, and grow your active user base?

To get the ball rolling, you could have the runner pick a goal for running and track the different weather conditions. The app could ask if the run was good or not — and then you can start to build custom schedules based on historical data," Steven said.

Steven suggested looking beyond features like times, trails, and personal bests that require data to work well. Instead, he said there could be a use case where an app leverages weather and historical run information to guide the end-user.

Using this data, the app could tell a runner that today is the best day for sprint training or long distance training," Steven said.

Telling a story with data

Data is a critical differentiator in most products today, but leveraging that differentiation to make a sale takes telling a story with the data. Continuing with the running app example, using historical running and weather data, the app could ask the runner to pick various goals and then provide feedback on how they felt and how the weather impacted them.

The value proposition could be that the app is creating a custom schedule built off historical runs and external factors that impact your feeling, efficiency, and speed for those runs," Steven said.

Understanding data sources, storage costs, and data processing methods can only get you so far — and they're critical questions that need to be answered as you work through customer discovery through to your first alpha or beta customers.

From a big data perspective, you're trying to collect as much as you can, and then you narrow and refine. My perspective on big data is about understanding what's important to you and to your customer," Steven said.

More data, bigger risks

Data can be compelling, but it can also be scary.

Being transparent about what you’re collecting, how you’re using it and what it means for your customer is critically important. What you can do with data can be impressive or shocking.

If you can build elements into your product that speaks to some of your long term goals, what you really want to do with the data, can turn a shock of ‘how do they know this about me’ into a feature that’s exciting and anticipated before it’s even released.