So Is Everyone a Software Company Now or What?

I think that agencies are starting to adjust their commercial offering into something that charges clients for access to software, as well as for access to intelligent service-delivering humans.

I also think that this represents a fundamental change in how agencies package and monetise the one thing that’s theirs to sell: their knowledge.

I’m not sure if this represents something paradigmatic in how the agency business shall operate from henceforth and for evermore; it could just be that my opinion is formed out of speaking into a vacuum-chamber of consultancies that see things in a similar way to each other, and are altering their approach accordingly.

On the other hand, it could all just be a simple counterswing reaction to AI’s ferocious potential to utterly consume everything that we previously relied upon for consistent services income, forcing our industry to find new ways in which to be creative lest we shrivel up and die.

For the sake of simplicity, I think we can assume that there are multiple originating reasons why it’s happening.

So what exactly is happening?

I’d like to briefly recount three scenarios. They are all real, all based on conversations I’ve had in the last six months, and are all anonymised to within an inch of their life as I want to preserve confidentiality.

1. Souped-up MMM A company specialising in Media Mix Modelling (MMM). Their consulting projects tended to be very long, with many of their projects dragging on due to delays in collecting data sources and getting them to talk to each other. To speed up this process, they developed software that assisted with the ingestion of messy data, transforming literal (I mean metaphorical) buckets of PDFs into a single CSV.

2. DIY Data Studio A performance agency with an analytics service line. They vibe-coded their own Data Studio replacement, with a skinnable frontend that could match clients’ own design preferences and branding. It included an “analytics-on-the-analytics” admin area that showed which dashboard elements were and were not being viewed or interacted with.

3. Intelligent Query Routing An analytics consultancy designed and implemented a semantic layer for an enterprise client. They built a routing model that automatically pipes queries to the most efficient model when users across the organisation query their internal AI tool, helping the client save money on token costs by routing simple questions to lower-cost models.

What do these have in common?

Well, they’re all smart solutions.

And they all address real, actual, not-made-up problems (and don’t forget that, at its most fundamental level, the entire business of consulting is all about solving problems of one kind or another.)

And yes, they all make extensive use of the accelerated potential of software development thanks to the vibe-coding, agent swarms, and CLI superpowers enabled by AI.

I think there’s an interesting contextual factor that they all have in common, too.

Agency software is supposed to be internal use only!

I would guess that about 92% of analytics service businesses have, at some stage or another, invested non-billable time in the development of software of some kind.

It might take the form of a Google Sheets script that cleans up data imports so that data can be more easily ingested into a client’s environment, for instance.

Or it might be an auditing tool that lets you easily see the difference between two Google Tag Manager containers without needing to be able to sight-read towers of JSON from 20 paces.

In those sorts of situations, the stuff that gets developed is for internal use only. It’s not for public consumption.

I suppose people might allude to it in client pitches (“…and of course, our internal tooling enables us to do things faster than the next group of people you’re about to have in this room who will tell you the exact same thing…”)

These tools are great. They help you deliver your services more efficiently, decoupling the value you generate from the raw person-hours of labour you engage.

They have a competitive edge, letting you lower your pricing and/or increase your margin, winning new clients and offering better service to current clients.

Greater efficiency is how agencies get to punch above their own weight: if your tools can handle the complex sprawl of enterprise analytics setups, you get “promoted to the top flight” and start working on bigger projects for bigger clients for bigger bucks.

All of these are valid; but the three examples above are different.

How?

Well, clearly -- if the client is getting access to the tool, it’s no longer internal.

Internal-only software is used by the agency to help them deliver the service. The client doesn’t even need to know that the software even exists. The thing that the client pays for (the product) is the completed service.

Once the software becomes external-use, and the client can see it, the thing the client’s paying for is the software itself.

That’s significant, but it’s not the entire story of what’s changing.

If we think back to a pre-AI, services-only agency model, the situation we had in those simpler, idyllic times would see an agency taking their intellectual property and know-how, and package that up as a Service.

Knowledge gets bundled and then delivered, with an agreed output, and the client pays the bill.

So you could crudely outline it like this:

Knowledge → Service → Invoice

We can think of that model as being Knowledge As A Service.

S + S = ?

Now, of course, things have changed and we’re Knowledge As Software Plus A Service.

The throughput of knowledge to service is still present, but now we see something else happen too.

The MMM tool for instance? The agency used their knowledge to develop something that would expedite the delivery of their service. It makes the journey from Quote to Invoice more predictable, more forecastable, quicker. All good things!

How about the Data Studio replacement? The service provider wrapped their knowledge into a tool that helps them work faster, and locks-in a client to using their platform on an ongoing basis. They used their knowledge to make clients stickier… a very good thing!

And the LLM-routing semantic layer addon? While the most niche of the three, you can guess where I’m going with it: they packaged their Knowledge as a subscribable service that locks-in an enterprise client… an unquestionably good thing.

In each case, the consultancy’s moat is their knowledge, but it’s also how that knowledge is deployed.

They know how this works from a service delivery perspective, they can use AI tools to accelerate the development of software that assists the delivery of said service, and they monetise the combination.

We never had that combination when we were pure-services.

It’s not just me, right? This is a weird shift.

Perhaps we’ve left the days of Knowledge As A Service (ooh, can we call this type of business a KaaS?) and entering the days of Knowledge As Software Plus A Service (omg omg omg this abbreviates to KaSPaS, pronounced like the collective noun for a group of friendly ghosts!)

Another way to look at it: these days, spinning up a tool is easy. The software becomes a container for the same Knowledge that forms the core of your Service line.

Historically, agencies would develop tooling to use internally, and as we’ve seen this is good as it protects margin, drives stickiness, adds value to client relationships, and would also increase your business’ valuation in the event of any M&A event.

But today, the process doesn’t stop with tech that gets used internally. It’s allowed to mature, to extend to real-world client situations, and the same things it was solving before (margin, stickiness, adding value) get solved much more effectively.

So does that mean everyone is a software company?

I think so. Well, probably.

It’s reasonable to assume that service-led businesses will become increasingly adept in developing tooling that sits alongside their services, with access offered under contract to paying clients.

That doesn’t mean that I think that any agency that doesn’t do this is doomed to extinction. I don’t doubt that the future has plenty of room for everyone; if you’re good, then you’ll survive.

If you’re the founder of a consultancy, or you’re otherwise responsible in some way for delivery of analytics services (which I assume you are, otherwise why on earth would you be reading this), perhaps you’re feeling a vague, over-the-shoulder feeling that you need to change something in your business.

Perhaps this might be it?

But hey, enough of my yakkin’. Stay well until next time!

---

Thanks for reading. No AI was used in the creation of the words you just read.

I do this for a living. I work with agencies and consultancies around the world, providing the support I wish I’d had throughout my journey of starting, scaling, and selling an agency.

If you think you could benefit from what I know how to do, I welcome conversation: send me an email.

On this site you’ll also find information about the Analytics Consultants Circle, a free monthly peer-group call that I moderate, where solo and small-team consultants share battle stories from the field, and we talk about the challenges you face when you grow and scale a business.

The last movie I watched: The Commitments (1991).