SharpGrid helps clients increase sales, market share and solve business challenges. But who is behind the entire data ecosystem that makes our products and services viable? Meet our Senior Data Analyst, Tomáš Kovařík, known here as Kovo, with whom we take a look behind the scenes of SharpGrid's data world today.
The key for me was studying database systems at university, after which I started working as a BI consultant. I'm more focused on the technical side of things, but at the same time I've always been attracted to solving business challenges with data and was looking for a place where I could apply both. That's when an offer came from BizMachine (which SharpGrid later spun off from), which was dedicated to doing exactly what I enjoy - working with interesting data in an area where you can get to solve a really diverse set of problems.
I joined BizMachine before SharpGrid was formed, but I was in the on-trade section from the beginning, so my subsequent transition to SharpGrid was seamless. The first version of the data output, which is our main product today, was created a few months before I joined, so I've been involved almost from the beginning. You could say I was almost at the birth of the whole idea.
Our data is unique in one way. Everything we do is closely tied to the real, tangible world out there. I'm fascinated by what can be learned about it from data sources like Facebook, Tripadvisor, or various websites. We interact with social media and the web every day, but it wasn't until SharpGrid that I fully understood how widely the data could be used. What I enjoy most about my job is figuring out how to work with these huge volumes of information, expanding it and finding new uses for it.
I found it interesting to process data from food delivery platforms such as Bolt, Wolt or Delivery Hero. It's an area that's been growing very fast lately, and there's also a whole new type of food outlets emerging: the so-called ghost kitchens, or places that you can't physically visit and only provide delivery. If I had to pick something more technical, I really enjoy processing data using tools in the cloud. We have a whole team dedicated to it now and it's broadening our horizons.
We mainly help them identify their relevant market, for example all the restaurants or bars in a given country. Our clients typically have information about the places their sales reps visit, but that doesn't represent the whole market, it's just a slice and using it can ultimately lead to wrong business decisions.
The accuracy and consistency-in-time of manually collected data used to be dependent on the skills and time capabilities of sales reps. In contrast, we provide our clients with a complete picture of the market based on online data and a digital footprint of individual outlets. This is much more accurate, saves time and frees up the hands of the sales reps and their managers.
Another area is selecting specific segments based on what the client needs at the time. For example, analyzing or picking out specific types of businesses that they want to target, such as premium restaurants, nightclubs, or trendy cafes. Our data is so comprehensive that we can do an analysis over almost any segment or part of the market, even tailored to individual client’s needs.
Based on the data, we are also able to tell clients which establishments are heavily dependent on summer/winter seasonality, what's on their menu, what their price point is, or what their potential is for selling certain products or product types. And of course, a lot more. For example, our Outlet Census product has over 40 different indicators that key people in the company can use to decide where to send salespeople, where the best chance of closing a deal will be, or where there are some unexpected business opportunities.
The relational and non-relational databases we use to store data are the mainstay. We work with MSSQL, PostgreSQL or MongoDB on a daily basis. We mainly use Python for data mining and processing, modelling, and data analytics. We also implement data transformations in Python using Spark cluster and other cloud tools in Google Cloud Platform and Azure Cloud. We use Apache Airflow for component orchestration and processing.
The big challenge is to maintain high quality as we work with data that is constantly changing and evolving. For example, it's important for our customers to have the most accurate information about the outlets that have just started operating and those that have recently closed down, which is not easy to track. But we can do it.
It's similar to linking our data with the client's data. Clients typically have their data in a CRM system, and in order to make the best use of it, we need to connect the two. Often the same outlet may have different names, a different address or may no longer exist in one system or the other. Automatic data linking is simply not easy. But again, we can do it.
As a senior data analyst, I'm involved in most parts of our data ecosystem. Lately, it's been mostly client specific projects and standardizing different parts of our product.
My typical day starts with a morning standup with the rest of the team. Afterwards, my colleagues and I solve problems they've encountered that they need help with. This might be adjustments to data transformations or revisions to our deliverables for clients. I'm also in touch with the commercial team and we talk about our clients' needs, their feedback and how it might impact our products and deliverables. Typically, I also deal with planning the team's work, preparing outputs for clients based on their requirements or helping with the development of SharpGrid's Outlet Census and Market Meter products.
Ideally, I should just watch everything automatically work itself out (laughs). Technically, we want to have the basic steps automated and use our time more efficiently to tackle the really challenging and interesting tasks, like extracting new insights from data or expanding the indicators we provide to clients so they can better manage their business, plan strategy, and manage teams accordingly. That's where our biggest added value lies, and I also just enjoy it. In addition, as we expand into more countries, there are bound to be new challenges that we haven't yet addressed.
The first one is creativity and problem-solving skills, because every project is specific in some way and you always need to look for new ways and solutions. This applies to both our product development work and client-specific work.
The second is a pro-client approach. It's not enough to just deliver data to a client and not worry about whether they used it at all and got something out of it. Our work is heavily influenced by what the client needs and for what means will they use our data.
And the third is teamwork. The whole team is involved in the products, development, and deliverables, so it's important to have feedback from colleagues, to learn from each other and clients, and to share our experiences and insights with others.