Changing E-Commerce Software Packages - Part 4: 14 Day Trial with Selro

This is the next part in a short series of posts about the process of changing the core software I run my business on. Part 1 was about clearing down existing workload and the overview of the project, Part 2 covered exporting the data from the existing software package, and Part 3 looked at key requirements for the new software and a shortlist of possible providers.

As before, it's mainly so help me document for myself what I've done, but done on Hive because it might help anyone else who is thinking of doing something similar.

Image by REDQUASAR from Pixabay

Overview

I picked Selro because it appears to tick all the boxes in terms of the requirements I outlined. Additionally they are a relatively local UK-based company.

They're a smaller company than Linnworks, our current provider, and reviews indicate that their support can be a bit inconsistent. But that seems to be the pattern with support wherever you go nowadays, with the advent of AI-generated responses seemingly making things worse, not better - a faster response time is useless if the information doesn't accurately answer the question !

Initial Sign Up and First Minor Hurdle

The initial sign-up was very straightforward; pretty much the same as signing up for an account with any other online software package. I was also able to book a short onboarding call a day or two later.

While I was waiting for the onboarding call I looked through the online documentation.

One thing was immediately clear, which is that the setup process in Selro is designed to follow a very specific path. The first step is to integrate your selling channel with the "best" product data and import the product catalogue. This would become your Master Channel.

This was the first hurdle. Not all of our products are sold on all channels, and the details (descriptions, titles etc) vary from channel to channel. I prefer to import by csv from a master spreadsheet, which is far more accurate. Their method would work well for a new business or one moving to this kind of software for the first time, but isn't so good for established businesses migrating from another package.

Additionally, our "best" channel is our website. Our webmaster had advised us that he was tied up with another project for a few days and it might be best if he was around to be "on tap" should there be any problems. We're only a micro-business, so he was being very reasonable, I wouldn't expect him to drop another customer's pre-planned project if I messed up !

Initial Channel Integrations

With the website off the menu for a few days, we did the onboarding call, which was helpful although I felt that at this stage I wasn't really sure which questions I should be asking.

Then I imported the product data csv files I'd prepared. This was a bit of a slow process; trying to upload all 10,000 products at once froze the system, so I did it in batches of about 750. Not a problem, this should be a one-off exercise anyway.

After this, we did the eBay and Amazon integrations. While they were very straightforward, I did notice that the online documentation was subtly different to what I was seeing on screen. Probably down to updates not being added to the documentation. Again, this appears to be the norm nowadays (even PayPal did it to us last week !)

This was also when I discovered that there was no easily accessed new customer dedicated support, I had to submit tickets just like any other customer. The average turnaround time was 12-24 hours.

This is a concern; under normal operational conditions, 24 hour support is usually fine, but not if there was a business-critical outage or urgent problem. When learning a new piece of software, it's frustrating to have to wait that long for the kind of questions which usually had a very simple solution.

Royal Mail & Website Integration

Despite these minor niggles, I was generally happy with the software. I could see orders flowing through; we were still doing the actual processing through Linnworks, but Selro was picking up the changes.

Moving on, it was time to start integrating shipping methods. This initially meant a Royal Mail integration. I'd been dreading it a bit because anything involving Royal Mail always seems to be over-complex.

In this case, the biggest issues were that Selro's documentation had gaps in it, and their index of Royal Mail services wasn't a direct match for the service codes we get from Royal Mail. It required quite a lot of head-scratching and was far from intuitive.

This was also the point where our webmaster became available for us. After he did a full backup, we proceeded with the Woo Commerce integration. The actual setup was a very straightforward API connection, and I could see orders starting to flow through.

Show Stoppers !

I soon spotted that there was an issue with the product catalogue. Selro had decided to treat eBay (the first channel we'd integrated) as the master channel, so wouldn't let website data over-write anything that had been imported from eBay. A big pain, but with this still being officially a free trial it could be solved by clearing the data and following their process from scratch.

Then orders stopped being pulled into Selro.

Our webmaster has built us some good security, and it includes automatically locking out bots dialling into our website to carry out mass operations (anything from DDoS to data scraping for competitors) unless they are whitelisted.

The Selro API had come to our website from a US-based IP address, and wasn't on the whitelist. I sent in a support ticket to Selro to ask for a list of IP addresses they used. It took them a full 24 hours to get back to me with a list.

The very next Selro IP address that dialled in was not on the whitelist they had given us.

After discussion with our webmaster, we identified that the issue is that they use AWS (Amazon Web Service) cloud servers, spinning new ones up as required. This meant their API could come in from dynamic IP addresses any time they needed more data processing capacity.

Aftermath

My webmaster and I agreed that manually whitelisting IP addresses on a daily basis was not practical. Opening the website up to unknown IP's to prevent the API failing any time a new IP address was used was too much of a risk. With Selro taking 24 hours to respond to what would have been a business critical issue, we decided it was time to terminate the trial.

I was very disappointed; the people at Selro were lovely to deal with, and the software has potential. It's not a bad package, it's just that it's not quite right for us.

I don't regard the process as a total failure; finding out if there are show-stoppers is the whole point of doing a free trial. It made us look hard at what we do, and I did quite a bit of data cleaning in the process.

Spending two weeks working on this on top of day-to-day operations was a significant challenge with an opportunity cost in terms of mundane stuff I just didn't get time to do.

For the time being, we have no choice but to stay with Linnworks. I've got other projects I need to work on for a while. But I'm sure in due course I'll be ready for the next trial, and I'm better equipped to identify pitfalls and problems faster.