Skip to main content

Why No One Should Ever Build Their Own Search

Peter Messana's headshoot
By Peter Messana - CEO

I do not write a lot about Searchspring as I made a pact with myself that these blogs would not be self promoting and would instead be about ecommerce and the ecommerce tech industry, but I am going to cross that line slightly here for the benefit of everyone. Before I jump into why you should never write your own search engine, I need to backup a little bit and tell a short story.

Back in 2005 I launched my first ecommerce site and built my first ecommerce search engine. I am using “I” because it was all me, we were a two person retail company and my job was to get a website up and running.

The first search I built was bad, very bad. So was the second and the third. Even as I got better each time, I just couldn’t keep up with the competitors. Search is difficult, very difficult. So, I did as so many others did at the time and reached out to Google to power my search. I threw in the towel.

Out of everything on the site, search was the only piece where we used a third-party provider, everything else we built ourselves. We didn’t have the luxury of a Shopify, BigCommerce, or Magento but it was clear to me that search was difficult and hugely important. Search converts 5-10x browse traffic. Shoppers who search are the highest intent customers on sites.

My experience from there weaves through a few different providers as Google exited the market, but the main point is that I completely understand ecommerce search and how damn difficult it is, and I am here to tell you, never try to build it yourself. I’ve seen the best in the business build their own and the economics and performance never work out.

Over the last 5 years I have been involved in two different search companies and my comment still stands, search is difficult, very difficult, but it is what we do. Every. Single. Day.

The problem with building your own is that you build it but it is never done, the total cost of ownership (TCO) is astronomical. The support and maintenance along with keeping up with new features means whatever your first year cost is will likely be the cost every year going forward.

Take a fairly decent sized company doing $100M in online revenue, they likely have a decent size ecommerce engineering team. If you want to build and maintain a search engine you are adding 2-3 dedicated people, minimum. They will build it and then be consumed with scaling it and adding features and integrations.

When you look at Searchspring, we have over 35 engineers (and growing) working to make search better, which means we are working day-in and day-out to keep up with the latest and greatest features and integrations that clients want. There is always the argument that if you build your own you only have to worry about the features you want, and we have to worry about features everyone wants. This is partially true, but the features we build are only ones that we know extend across the ecosystem, so when someone has a great idea that they want to add, we already have it built or can build it and then extend it.

So if you are sitting there thinking about building your own, shoot me an email. I will talk you off the ledge, and even if you aren’t using Searchspring I will still help you out.