Hacker News Reader

Comment by dhruvkar | Hacker News Reader

Construction is a big field, and there are many niche industries within it.

I work in natural stone. There are several different types of companies just in this niche. Usually the flow of product is:

quarries -> factories -> distributors (us) -> fabricators -> kitchen and bath shops -> end user

There are maybe 5-7 major pieces of software for the first four stages. "Under-served" is an understatement.

Similarly, there are 10s (if not hundreds) of other niches within construction, each with its own vertical. Tons of niche software is missing.

When I see another niche analytics/forms/site chatbot/automate social media startup, I slap myself to make sure I'm not crazy.

There's a whole ocean out there, but most software startup companies are running into each other in Depoe Bay.

View on HN

And all of it highly digitized with 3-4 generations of technical debt and path dependencies that needs to be dealt with before any mobile/saas solution will be anything but a waste of time/money for the organizations involved.

Where you see badly digitized enterprise organizations with high admin overhead you always find an hugely expensive mid/late 90ies windows PC centric client/server environment in place blocking nearly all attempts to migrate back to anything centralized enough to support an modern saas solution. i.e. the problem is rarely that they are undeserved but that they are stuck with an nearly good enough computer platform and too much complexity for anyone to attempt an big bang replacement.

i work in one of the niche's you mention and we are 110% digitized with about 5 different solutions in place for every task causing an ton of manual cleanup when they inevitable fail to talk to each other correctly.

>> problem is rarely that they are undeserved but that they are stuck with an nearly good enough computer platform and too much complexity

This is true for large, enterprisey companies. But I'm talking about companies in $7-15M range who have bookkeepers downloading 35 excels each month and then run their custom Excel process to enter into a basic ERP system, so they can send out emails to each of the 35 employees to collect receipts. That kind of thing is so prevalent and accepted, the biggest hurdle here is convincing that there is a better way rather than "hire another bookkeeper".

The problem is that it kind of isn't because there is no clear way for your app to work with the app their employee's use for taxes or the desktop apps used by their supplier/customers.

At the end of the day the 35 spreadsheets represent an workflow that your generic app will likely not automate, it's also not an empty market as every region have 5+ companies selling the magic silver bullet for mid-sized company bookkeeping, and all of them have an high degree of lock-in and cannot simply export it's data in an rigidly standardized format.

The experienced small company manager know that an experienced bookkeeper can make the administration spreedsheetwork be an low frustration experience, and that the previous software vendor could not deliver the same.

>> 35 spreadsheets represent an workflow that your generic app will likely not automate

This is part of knowing the niche. The "app" doesn't need to be generic, or an app at all. Add-ons for existing industry software, integrations between industry software and generic inventory management app, etc. are all ways to provide immense value.

>> not an empty market

Yes, but they are often flooded with generic apps. "Inventory management for SMBs" because there isn't anything more specific. Until a company is large (enterprisey) enough to bake their home-made solution.

>> experienced bookkeeper can make the administration spreedsheetwork be an low frustration experience Not always. Definitely not true for our last two bookkeeping hires in the last 6 years. Part of it will always be a slog and as the load gets higher, sometimes the right answer is "hire another bookkeeper".

In the "35 spreadsheets" scenario, I have found Quantrix to be the killer app. Very occasionally, Sharepoint and server side Excel is viable, but if you can import cleanly, Quantrix is almost there for claiming small proprietor mindshare. The beauty of anything that is easily understood as more robust logic, in a proprietors view, is the path that opens for you to work on the tricky bits they'd never let you touch otherwise.

>> Quantrix

Hadn't heard of it, I will check it out, thanks!

The problem is that you need to be niche specific all the way down to specific region and specific supply chain and customer base for your app not to be "generic" and then we are practically doing custom coding for an specific customer.

I think we're just arguing for the sake of it now.

There is space for both viewpoints.

>> Specific Region - the US.

>> Specific Supply Chain - consisting for 100s of companies

In my metro area alone (not in the top 20), there are hundreds of one type of company, not counting the others in the supply chain.

Catering to a niche like that, i.e. building software where the customer base < 1000 companies is still a potential multi-million dollar software business.

You are absolutely right, if you niche too far down, you'll get stuck customizing for each customer and that may not be worth it.

And there are a couple big exit paths. Autodesk, Trimble, Oracle, Nemetschek, RIB and for more specialized software InEight. If Procore stays independent, and goes from innovator to incumbent, they may start to absorb new combing competitors as well (like Instagram and whatsapp, were hedges against market share loss.) CMiC, IFS, Oracle Axonex and Trimble Viewpoint (including Spectrum) can't be ignored.

If you're looking for ideas, look for things those companies don't do, or largely do with legacy components to their suites. Up and comers include Revitzto, Katerra, Alice, BuildSafe, Rhumbix, BioSite

The other thing construction doesn't do is edi. Compared to logistics, all POs between firms accounting systems are pdfs and data entry. Interoperable accounting systems will be a huge paradigm shift, where blockchain could be more than a buzzword.

The real future in BIM is around things like Autodesk Plasma and Forge. https://www.aecmag.com/technology-mainmenu-35/1821-beyond-re...

>>The other thing construction doesn't do is edi

This was a struggle for us, and we decided not to pursue it. Trying to get our company, which is not in the logistics field, set up to receive EDI updates is a pain. Getting all freight forwarders to add you as a notifying party, paying for an EDI service or setting up your own. Getting approved by shipping lines that don't recognize you as part of the supply chain.

Instead I've opted for taking all freight forwarder PDFs and parsing them (with some ML) to build our own dashboard of what material is coming in.

the problem is having access to these niche fields and knowing enough to be able to create something useful.

agreed, a lot of it is niche knowledge that you could only gain by working in such a field or having close access to a company.

Consulting within a niche is one way to do this.

In the eighties, my best friend's father was into organic farming in the following manner : he'd squish the slugs he happened to find while walking through his crops and that was his pest control over the top of great land husbandry in the tenth generation. My friend taught himself turbo pascal then vb and the farm tax deduction provided a 286 of my unfulfilled dreams. The son was shortly trading in commodity supplies and even derivatives and taking delivery of GPS equipped combines. But I doubt he could have been successful had he not grown up with his father's wisdom.

I would say the biggest challenge of this kind of field is it is so big so no one actually know everything. The rule also varies between countries. For example, the construction rule in England may be different from the rule in the US

That's what's exciting for me. If I have two strengths related to startups, it's being able to write code and get up to speed in new fields quickly.

Interesting that you mention construction, since Procore just filed their S-1 and they seem to be focused on this industry.

Absolutely agree, construction is very big field. I work for software company targeting construction, the scale of data, operation is far beyond my expectations

Reply on news.ycombinator.com