A common requirement of a commercial website is for an online store locating tool. You know the sort of thing – “Find your nearest store”, “Branch locator”, “Store finder” etc. You pop in your postcode or address, and up pops the nearest stores in order of distance from your location.
For a while tools like Batchgeo (http://batchgeo.com/) offered a nice, easy, feature rich solution. Whilst not terribly customisable, it did allow you to easily upload data, geocode the addresses and embed the resulting map into your website with only minimal Batchgeo branding appearing on your site. This suited my client’s needs for some time…
However, it never quite worked the way the client wanted it to, and there was a slight bug when doing multiple searches – it would kind of get stuck on the old address and not zoom the map to the new location properly. The tipping point came however, when Google introduced a 2500 addresses per day limitation on the ‘geocoding requests’ it processed. Google’s tool underpins a vast majority of mapping software these days, so this new limitation had far reaching consequences.
There’s nothing wrong with Google’s fair use limitation, and they do offer a paid for service as well (https://developers.google.com/maps/documentation/business/), but Google’s limit hampered Batchgeo’s free tools quite considerably as well. The problem arises because Batchgeo attempts to upload and geocode all your data in one go. Because of Google’s daily limits – Batchgeo have now had to introduce a limit of 250 addresses on their free account, no where near the 3500 addresses my client requires.
Ok, so Batchgeo offer a paid for service as well, but why pay $99 per month for a service the client was never totally happy with? That would be a hard sell getting them to swallow that cost indefinitely.
The answer for us was to build a custom solution. The client could spec exactly how they wanted it to work, it would be 100% on-brand, and there’d be no on-going license fees to pay – so would actually work out a lot cheaper in the long run.
We looked at some examples of other “Store locators” online and combined the best features into our new tool. Some of the key features include;
– Option to detect user’s location automatically
– Display all the stores within a 25 mile radius (with the nearest first)
– Clearly display branch contact details and full address
– Option to drill down to view a map and directions to the store (pulled in from Google maps)
We built the system to automatically check for a CSV file on the webserver containing the address data (exported from the client’s CRM database). This is exported daily. If the system finds a new file it processes the data and updates the database.
The Google fair usage limitation was not a problem, as we were able to code the software to only re-geocode the addresses that had been modified since the previous upload. This way once the initial upload has been completed, only one or two addresses may require geocoding each time (eg. when new stores are added or existing stores change location). This speeds up the process and also vastly reduces the load on Google’s servers (the aim of their limitation in the first place).
Our client ultimately got a solution they were much happier with, built to their requirements, and automated so that they don’t even have to lift a finger to update it.
You can see the new store locator in action here: