Time to take the lead and offer my 16 million fellow citizen in the Netherlands an app for Smart Grocery Shopping? Or is this one going to be out of my league? Let’s understand the challenges as much as possible before we embark on a mission that is doomed to fail from the start.
Next to describing the Why as we did in our first post, it might also be a good idea to describe the Why not. Why didn’t we find our app? Why are the other parties potentially not having the same ambition or vision we have, or why don’t they succeed in achieving exactly that?
An important factor could be an elementary lack of up-to-date data on the important entities for our model. Shop location is not moving that quickly and for that well maintainable. I can easily imaging to be able to extract supermarket locations quite easily from a number of source, like shopping brand websites, chamber of commerce feeds, or sort alike. Product entities are however much more in flux, and certainly their price property is subject to daily change. Such an information stream should be coming from a retailer itself, period. And that is where the pain might be.
What’s in it for the retailer to provide a list of there most actual prices to a third party and allow them to compare prices with other retailers? For only 25% of them it will strengthen their market position, the budget stores. For 25% of them it will weaken their market position in comparison to the other retailers, the premier, more expensive ones. For the rest, it does not provide additional value to be on the list. Having said that, in other markets like electronics, cars and houses, shops do push out up-to-date info to big comparison sites. Those markets have matured into the model that ‘if you’re not on the list, people are not considering your shop for buying’. How to overcome the initial chicken-egg problem of a missing initial list meaning there is no value to be on it, because it is not there (tell me which one is the chicken?).
Data is everything
Above analysis feels quite plausible to me, and I think most established parties with the intent to compare prices over retailers are having a hard time getting their hands on the data. Their infrastructure is in place. Their information model is sufficient to capture the data needed and get the outcomes we envision. Technically, it is very easy to realize what we our looking for, and they’ve got a huge head start.
Getting the data from the right source is a communication, commercial and sales exercise, more than a technical one. You need to start sending emails and letters to retailers (or joined retailer organizations) to get hands on the data. Have lunches with board members during which you will sell your proposition and make them understand it is in their best own interest to participate. Access to a technical APIs to connect to is a potential outcome of that non-technical effort.
Am I a person that can pull that of? I’m rather a techie than a sales guy. Do I have contacts in that domain that I can leverage? Not really, or at least not that I’m aware of. (Note to self: check my LinkedIn contacts. There might be valuable first or second grade connections I could contact). In a project I did several years ago, I was lucky enough to quickly find another party that was aggregating all the information I needed, and was providing a connection to their API in exchange for traffic to their side. I only needed to create the client app. Is something like that realistically achievable for the grocery retailer domain also?
The key takeaway from this post for me is, that:
It does not make sense to start writing an app without having a proper high quality data stream available.
We can elaborate on the envisioned information model, user interface and even do some mock-up. But without having the data available, we won’t start implementation. And with that, I’ll end this post. Still enough open ends we can work on in next posts. Hope to see you back real soon.