Accounting system and accounting services core

Every business needs an accounting system, and most businesses need help with accounting.

This means every business in Norway can be our customer.

But we are not serving businesses with some “random” service; accounting is high trust. This means advice about adjacent systems is much more likely to be followed.

This opens up for something extremely valuable: Upselling.

How AI changes the system / SaaS game

In the past, it made sense that separate teams/companies build separate products, and that integrations (via APIs) were made when data needed to be exchanged between systems.

This has a few disadvantages:

• Higher cost
• More suppliers to deal with
• Network communication between systems is inherently unreliable vs. one system

Just to take one example: A business on Shopify might pay thousands a month just to get some sales/order data into their accounting system.

I think that the future is in systems that do everything a business needs, and that AI enables creating bigger and more integrated systems.

The future is less about connecting systems, and more about building a system that has everything.

How to succeed: Modules

Even with AI I think a lot of thought has to be put into how to modularise the application in such a way that you can easily extend the system to do more.

Starting with AI-enabled accounting at the core, interesting modules/upsells include

• POS, point of sale system
• E-com system
• Hosting and domains
• CRM
• Agreements and signing

The two interesting questions to ask:

• What do (almost) every company need?
• Which niche is big enough with companies that are less cost sensitive? (lawyers etc.)

A lot of fearmongering around SaaS exists, the idea that “now people can vibe coding something completely custom to my business, we don’t need to pay for this SaaS subscription”. This is stupid, most businesses are not that custom, and it is better to just pay for someone who has set up a proper structure with an API that lets agents automate your custom workflows, than to make a fully custom system.

I would say that is true 99% of the time and SaaS is more interesting than ever.

Finally, maybe the most important change in SaaS is that API no longer becomes an option. Agents doing work is the future, so building every system in such a way that they are agent-friendly is a no-brainer.

Where to capture value

I still think that a lot of value can be captured in both software and services.

I also want to briefly mention that there is significant value to be captured in payment processing as well, which I don’t really consider software or services. This will be relevant for both POS and E-com systems.

the merchant/transaction side (which is dominated by, but not limited to, the payment-processing cut) 
is about three-quarters of Shopify's revenue, 
and the recurring subscription fees merchants pay for their plans is about one-quarter.

Software

Tripletex was doing 500 million NOK EBITDA in 2024. Other systems are doing similar numbers, even though I think there will be some competition here, I don’t think AI changes this significantly.

If you can have even bigger and integrated systems, one could imagine ReAI doing over 1 billion EBITDA eventually.

Services

The opportunities here are endless. Starting with accounting services is very natural, but I think expanding to consultancy in multiple fields, especially software, can be very interesting.

People are willing to pay a significant premium to outsource hard tasks.

Selling software and eating it (dogfooding)

All software we build should be heavily used by ourselves.

That is the main value of Aiak. We can capture a lot of value, but it also makes it much easier to improve the system.

That creates a fast feedback loop:

• We see the manual work directly.
• We build the product around actual accounting workflows.
• We test it on real clients.

In less than a year of building, ReAI has already reached a level where I believe we have a clear path to becoming a major accounting platform.

Acquisitions vs. organic growth

Acquisitions should mainly be done to either:

1. Get the right people into our team
2. Grow customer base

Very few businesses have a moat that is worth buying and usually there is a ton of legacy to deal with.

Organic growth and a big sales department is critical for success. And I think it should be a higher priority than acquisitions

3-5 year plan

Top priority:

1. Succeed in Norway as an accounting system and (group of) accounting offices
2. Make adjacent systems in lucrative or big niches and upsell to our customers

Only after we “succeed” in Norway (success definition TBD):

1. Expand into Sweden/Denmark
2. Look into improving profit margin by going into highly regulated industries and own more of the “complicated” stack

I already have CPR number and experience with Danish businesses, so that might be the easiest place to start. I also think it can be possible to expand to Sweden as well in parallel.

When we expand to other countries, I think it makes sense to reuse more than Visma does.

Instead of having Tripletex, E-conomic, and Bokio as 3 different systems for 3 different countries, I think one system for all countries makes more sense.

Discussions / questions

Why can’t other systems replicate our differentiation?

Pricing:

The legacy systems have too much to lose to slash pricing, I think it is more likely that they will continue to increase pricing.

Our pricing is currently maybe too aggressive, we don’t necessarily need to be up to 10X cheaper (holding). I think that we can be on average be 3X cheaper on the accounting core, and then if we’re able to upsell other modules that they would usually pay for as separate products at “market price”. Remember that if they had a CRM, they would usually need to pay for that separately + an “integration tax”.

Standardisation (of accounts):

I think it is much harder to do in a legacy software project than one might think, AI doesn’t solve this easily. It also takes some “guts” to do it, and big orgs often lack that.

AI automation:

AI automation is mostly about having an open API that is easy to use for agents. The problem legacy systems face is that if they open up for agents to easily and makes CRUD(create, read, update, delete) available for the agents. Then the cost of switching systems also significantly decreases.

Most systems support SAF-T, but it is still a lot of unnecessary manual work exporting everything and re-creating in another accounting system. That disappears with a proper API easily used by agents.

Some other systems will do this well, but to provide a better product they would need to be vertically integrated and provide an agent friendly API.

Where is the moat?

First, I would say that most businesses succeed without a moat. I think our moat in practice will be that we go very broad and that we become a “one stop shop” for what most businesses need.

Can the owners of foundation models take the application layer as well?

I don’t really think that “claude for business” will ever be able to integrate with Skatteetaten and Altinn. These are too niche to ever be interesting. Thus the owners of companies anyways need the accounting system, and why not use that for everything. It is the one system that will be a “must”.

Same thing applies to things like electronic invoicing, you need to be certified by Peppol or pay expensive third-party.

Big system Vs. specialised + agent orchestration

I think this falls into the category of “Interesting for IT guys, not interesting for normal people”. Most people don’t want to deal with agents or anything remotely related to configuring anything technical.

Unless the specialised systems are all built by the same company, I think it will be a worse solution.

“Saas is dead” from a pricing perspective

I think most of the costs of succeeding with Saas is still there in terms, and the biggest cost is marketing/sales. Thus even though some companies like ours has started out very aggressively on pricing, I don’t think there will be a “race to the bottom” unless it can make economical sense for multiple actors that can hold prices low for a long amount of time. That could only be possible if these companies upsell other services, otherwise it is no long-term business for them.

What will be the differentiation against Microsoft, Odoo etc.

I think the global players will have a hard time selling and making the best system for local businesses. The market will not be big enough for them to bother prioritising it.

We will not necessarily be differentiated on the “idea” level, but in practice I think we can build a significantly better product.

Why should accountant offices use ReAI if we compete?

This is similar to D2C Vs. B2B. Why should you sell on Amazon/XXL if they make your product and sell it cheaper.

Right now my gut feeling tells me that our main focus should be owning accounting offices and taking as much as possible market share here.

At the same time I think accounting services is so broad and sometimes also local, so it makes sense to allow “external” accounting offices to also use our platforms with clients.

There will definitely be friction, but hopefully our system is so much faster to use for accounting offices that it becomes a no-brainer to use it despite the competition.

IT managed service providers

I think it makes sense to go into this segment as competitors, if we imagine ReAI as a group of companies all built around the accounting system at the core, then I don’t think it is ideal if they need another tech partner to get stuff done.

Related to this I’m especially interested in the idea of becoming a registrar both as a leadgen and to faciliate helping clients setup domains.

Especially for .no domains the cost is very low. Not 100% sure about this idea, but I think it can be an interesting long-term bet.

How much of a boost does agentic accounting give

I think it is an absolute minimum of 2X more clients that one accountant can do. Agentic accounting also makes it possible for lower-level employees to handle more complicated cases without a senior authorised accountant helping.

I say this based on using agents to do accounting for clients myself. I have no concrete KPIs I measure, but I can obviously see that it goes in the same direction as coding. Things that take weeks, start taking days, and hopefully soon, hours.

Note that in Aiak we only have one authorised accountant that does controls mainly, the day to day accounting does not need an authorised accountant, that would be “overkill”.

Embedded payments

Here I’m in talk with Adyen. Based on initial pricing I think there is significant money to be made here.

There are primarily 3 targets here for a “ReAI pay” product (similar to Shopify payments)

• Pay by link
• Pay in POS
• Pay in checkout (E-com)

How capital should be used

I think external capital should be used primarily for:

1. Aggressive marketing and sales
2. Acquisitions
3. Funding new “company types” in the group, for example an auditor or IT MSP firm

Product development should easily be covered by operating revenue unless we want to go really broad, really fast.

I think immediately ramping up our sales team to 5-10 people makes sense if we take external funding. While bootstrapping requires a more gradual increase.

If we define success as 30 000 businesses using ReAI and 5000 of these using Aiak or the other accounting firms in the group, then I think we can achieve this in 3-5 years “easily” with 100 million in funding.

Example yearly revenue:

30k software clients. 5000 per year for software
5000 accounting clients. 25k per year avg
30000*5000+5000*25000 = 275 million NOK

I think we can even do this without external funding.

Poweroffice was sold to Visma for 2.3 billion only for software.

While this example also includes services.

An investor putting in 100 million for 25% of ReAI group, would have seen the investment increase value multiply by 5.75.

5.75 * 100000000 = 2300000000*0.25

So my current goal: Build something as valuable as what Pogo had at the point they were purchased in 3-5 years. But with a better foundation, and more room for growth.

Regarding cap table and investors I want as few as possible. I don’t think we need shares for employees or small investors that leads to “noise”.

The amount to raise can vary, it all depends upon the total deal. An investment of 50 million, 100, 500. Everything can be interesting as long as it is above “what I can easily loan + fund myself”.

Investor value outside money

I think the main value here can be help entering niches that are hard to enter without some “big money” behind you, I think the lawyer world might be one such example. But since I have not tried I don’t really know.

Networking, PR etc. can also be valuable.