Ask HN: Should we bring software dev in-house?

I'm a logistics consultant that specialises on the software side of things!

What kind of logistics does your company do? (Transport, Warehousing or both?)

Will depend a lot on your functional requirements, but I would say that unless you are doing something particularly unique, there are probably off-the-shelf products that do what you are looking for, and will probably be cheaper and more stable than a 3+ person dev team in house.

I don't know what your requirements are, but if you are in any way a somewhat normal logistics provider, what you are looking to do will quite closely match an existing software package out there on the market (or more likely, multiple packages). Just because you have package which is a poor functional match at the moment doesn't mean there isn't one that better meets your requirements!

In my experience home-grown systems give you exactly what you want in the short-term, and then come with massive limitations as you try to grow/scale them (i.e. if you are on the 3PL side of things, if you get a new client and have a good WMS you can probably on-board them purely with config without having to write code, despite them having some new/unexpected requirements), and the 'new' home grown system today becomes a legacy nightmare in the future.

Plus home-grown logistics software often misses some critical component that makes warehouses function well (e.g. I have come across many that don't have hard allocation, and then find that they have pickface shortage issues that are hard to resolve!). Unless you are closely copying how other software works in this space, you will probably fall into pitfalls that are already solved.

Assuming it is a WMS and you are a 3PL (my best guess from your description!), personally I think the best thing to do is get a good 'off-the-shelf' WMS and then dedicate your engineering efforts into the more customer facing side of things (e.g. customer portals) where you can actually show differentiation with your competitors. No point reinventing something that everyone else already has!

If you are a 3PL on the transport side, there are also great options that cover 'business as usual' and you can again push some development effort towards the customer side.

For logistics businesses, having software which is industry-standard and has a large support base is a bigger sell to a prospective customer than having your own 'great' homebrew software, but that's just my two cents.

Slightly boring answer.

Having been in that situation and developed SW in house as well as hired third-party firms to do so, based on my experience I would go with in-house if 1) you have very specific requirements that need to be well understood and implemented, and which may require close collaboration with the teams who are actually using the SW; 2) the SW in question is mission critical; 3) you want to be able to make changes/adjustments fairly quickly without the overhead of administrative/contractual dealings with a third party (i.e., whether it fits under existing SOW etc.).

If you do bring it in house look for very senior folks, ideally ex-founders who’ve built a product from scratch. You want someone who can help figure out requirements, feasibility, and write code. Pragmatic engineers are always the best.

happy to chat about this if helpful, email is in my profile

I have helped a company do something like this in the past. My advice is go for it, but understand that there's going to be some landmines.

The biggest benefit is that you likely have a stronger understanding of your needs than any third-party, and if there are other businesses in your niche, it provides you a pathway to build a revenue stream selling your custom solution to your competitors to capture part of their revenue. If you think about it from the perspective of value capture, it means even if you lose a deal on the front end to a competitor, you are capturing a percentage of it on the backend because it gets your competitors reliant on your platform. To do that though without running afoul of antitrust, you should try to set it up so in-house development of the platform operates relatively independently of the larger business (the legal term is "chinese wall").

As an outcome, you can actually generate a new revenue stream that's profitable while correctly serving the needs of your business.

So, base your decision only two factors:

1. Budget (perform a thorough cost estimate before making any changes)

2. Quality of service expectations

The risk part of this is that your company has never had a software team before. This means you have nothing to build upon. The level of risk here is tempered by the flexibility within your available budget, which means how much cost are you willing to absorb outside of plan if this goes to shit.

Now, set your expectations right. Software teams never ever make money. They only cost money. Will moving dev in house cost less money than the current solution? No, at least not at first, but you have to build from nothing. Moving dev in house will allow future scale the outside party does not provide.

Secondly, absolutely forget technology. Don't fucking go there and if you do you have already failed. Think only in terms of leadership. What that means for this effort:

1. Hire a leader who can perform risk analysis, measure things, work within a budget, is super assertive, and communicates brilliantly both in person and in writing. The assertive part is important because opinionated technicians/developers will attempt to drive the plan. Don't ever let that happen. A real leader will own this in both failure and reward.

2. Set realistic goals and accomplish those goals. Don't do anything else. Once the first goals are achieved you will have the foundation to do other things. Every distraction wastes more of your budget.

3. Form, in writing, policies that you are willing to fire people for violating. These should enshrine conduct, ethics, and appropriate standards for quality of service.

4. Do not (ABSOLUTELY NOT) think about this in terms of a startup. You are an established company already generating revenue/profit with margins. Proceed with a well planned vision always accounting for risks and make adjustments to the plan very carefully. Distractions and deviations will cost more money.

5. Finally, transition from the current solution to the in house software team carefully and progressively. This will cost more upfront, but dramatically less in the near term due to lower risk.

Good luck.