IT Service Sector as an Investment

Stocks
, , , ,
1474

A representative from Agion came to lead an AI workshop for us in the autumn. He boasted that the State Treasury (Valtiokonttori) had chosen Agion, a “two-person” team, and their agent farm in a tender over legion-sized challengers. Throughout the autumn and winter, I’ve occasionally tried to find this from official sources, without success given the epic search functions of the Hilma website.

Now, however, I found this short news item:

Or this:

I don’t know if Agion is actually alone in that and how many people are developing it, but if the claim about a two-person strike team plus an agent farm was even remotely true, it makes one think about the numerous rows of employees in IT service companies.

It would be interesting to hear comments from people in the industry. Is this common business reality, or is Agion one of the first examples of new small disruptors in the field?

12 Likes
1475

Wrong site. That €28k procurement by the State Treasury from Agion can be found on the tutkihankintoja.fi service: Tutki hankintoja. A procurement that small likely went through as a direct award or a mini-tender and therefore doesn’t appear in Hilma.

My understanding from the links above and from knowing Agion’s founder Alasaarela is that they aren’t actually disrupting IT consulting, but rather promoting the productivity of knowledge workers in public administration and expert organizations with their agent farm solutions. Of course, it’s a short leap to their own consulting deliveries indeed succeeding with a fraction of the usual human workforce.

As for these agent farms, they are definitely not BAU for these heavy-duty old (IT) consulting firms, but new AI-era players in Finland and globally have built their companies around AI from the ground up. Building an agent farm isn’t even a massive feat these days (platforms like n8n and the like take care of that), but a human-driven business model and culture are not easy to change in these established consulting houses.

Agion’s Alasaarela’s appearance on the Puheenaihe podcast is quite interesting to watch if you’re interested in agent farms more broadly:

13 Likes
1476

Digia’s results, to be published this afternoon and kicking off the earnings season for the Nordic IT services sector, will be very interesting:

  • The company is a significant player in the Finnish private and public sectors, which makes the outlook particularly interesting.
  • Profitability is expected to be very good, as no negative profit warning has been issued. Generally, the company’s relative performance during the weak cycle has been the best in the sector. Are the conditions in place for this good performance to continue?
  • The company appears to be one of the sector’s leading AI players (it’s very early to assess the leading players). Is this starting to show in the figures? Future winners are still very difficult to predict.
  • The company will likely publish a new strategy and targets for 2026-2028, as the previous period has come to an end.

@Joni_Gronqvist will comment on the results live in the earnings stream

Preview comment

Globally, the aversion to software companies has also been reflected in IT service companies. Here are the NTM EV/EBIT valuations for major global players Accenture, CGI, Globant, and Endava (NTM= Next Twelve Months). Historically, these companies have enjoyed high valuation levels (excluding CGI), but now only Accenture still reaches above the EV/EBIT 10x level.

image
image3268×874 420 KB

image
image757×477 36.6 KB

13 Likes
1477

What is this based on? But yes, the result is very interesting.

4 Likes
1478

Note: it seems that way. In my opinion, it is currently impossible to say for sure who is leading and who isn’t. However, based on the company’s communication, my own gut feeling, and this comment from the Q3 report stating that 15% of sales were AI solutions measured by contract value, I would say that the company is at least not significantly behind other players in the sector. Of course, one cannot know what exactly is counted as an AI solution here, but historically Digia has been quite conservative in its communication.

image
image1242×289 57 KB

13 Likes
1479

The AI implementations on Digia’s website aren’t really convincing. It’s hard to see them as a pioneer based on those references.

1480

Which companies do you think are the pioneers in AI?

As I also said, at this point it is very difficult to say who is leading and who is not. Situations will also change rapidly in the future.

5 Likes
1481

Franco, that’s a marathon where the first 2 kilometers are behind us. Everyone started from the same line, everyone had the same shoes and gear, and roughly equally strong training backgrounds. It’s indeed quite difficult to predict a winner. AI is the most equalizing technological revolution there has ever been.

10 Likes
1482

My view is that the winner of the AI race is the customer. Every IT firm will leverage AI in its development. More can be achieved with less. Ultimately, competition will drive the benefits to the customers.

11 Likes
1483

In my view, the question is almost structurally flawed in the consultancy scene. Consultants solve already known and well-defined problems; they don’t innovate the technological solutions of the future. Being a pioneer in AI is just cotton candy-flavored visionary fluff, not actual technological expertise. In practice, it’s the ability to get customers to buy pretty much the same services as before, which then have OpenAI and Nvidia logos bolted onto them. Both parties can then boast about their pioneering status. In the words of Linus Torvalds, 90% marketing, 10% technology.

The pioneers come from somewhere else entirely, and this industry follows behind. The challenge is partly structural for the Finnish consultancy sector.

Speaking of this marketing, Siili has gained good visibility in this regard, and it might even be reflected in their revenue. I haven’t seen any signs of special expertise.

8 Likes
1484

I largely agree with this, Harri, as I also tried to explain in my short comments :smiley:

From an equity investor’s perspective, what makes the situation very interesting is that for many consulting firms and software companies, the stock market is practically pricing in a run off a cliff over the coming years, which is why it’s essential to consider who will stay in the race and who won’t. Or alternatively, there are historically good buying opportunities available right now.

@exai This is a very relevant perspective. I believe that the bar set by clients for consulting firms will be higher than before, as the main driver of customer demand (in the big picture) is no longer a massive shortage of talent. At least in the coming years, the fact that the best developers are many times more efficient than average developers will very likely still hold true. This would support value creation for clients in those consulting firms that are able to recruit and retain these top talents. AI could, of course, theoretically level the playing field among developers, but at least for now, it seems that developers who are best at utilizing it have instead widened the gap to the average worker.

9 Likes
1485

The talk about being a pioneer is probably mostly marketing. I invested in Digia because I see it has good potential to be a “preferred supplier” for larger players like the public administration, who they turn to when they want to implement AI in their organization. This is one of the few areas where any development money is available or can be found in the public sector. A reference like Traficom is good here, because Traficom itself aims to profile as an AI pioneer among government agencies and also now has a significant role as the authority supervising AI.

9 Likes
1486

The pioneers are those who build production-ready systems with a completely new workflow, for example, by truly leveraging spec-driven development.

1487

Good point! Let’s level every company’s valuation to the same figure of 100, and start the competition all over again :smiley:

1488

Fear of AI Weighs on Software Companies – ”Micro-Hysteria”

Investors are selling software stocks at a record pace as new AI tools threaten to disrupt the business models of SaaS companies.

Tech Executives Try to Calm the Markets

Leaders of major technology companies have attempted to curb the panic. Nvidia CEO Jensen Huang called the idea of replacing the software industry with AI “the most illogical thing in the world.” According to him, AI will use and improve existing software tools, not replace them.

Arm Holdings CEO Rene Haas spoke to the Financial Times about “micro-hysteria,” noting that the corporate use of AI is still in its early stages and not yet massively disruptive.

ServiceNow CEO Bill McDermott said during an earnings announcement that market concerns are unfounded. According to him, the company’s products function as a “semantic layer that makes AI ubiquitous across enterprises.” Meanwhile, former DocuSign CEO Dan Springer estimated that he has yet to see an AI product capable of replacing what ServiceNow offers.

Salesforce CEO Marc Benioff has also emphasized the company’s competitive advantage: customer data and existing customer relationships create a moat that, according to him, AI cannot easily breach.

Analysts Split into Two Camps

Market commentators are clearly divided. J.P. Morgan’s Head of US Enterprise Software Research Mark Murphy called it an “illogical leap” to conclude from Claude Cowork add-ons that every company would build customized products to replace the entire enterprise software stack.

His colleague Toby Ogg described the situation more succinctly: the software sector is not only being held guilty before evidence is gathered, but has already been convicted before the trial.

A more cautious perspective was offered by Thomas Shipp, Head of Equity Research at LPL Financial. According to him, AI increases competition and pricing pressure while narrowing competitive advantages, making the valuation of software companies more difficult than before.

Consulting firm Constellation Research estimates that this is not a death blow for the software industry, but rather a concern that AI could weigh on company margins and limit pricing power. Futurum Group analyst Rolf Bulk told CNBC that AI will likely eat into some SaaS revenue, but companies providing critical enterprise applications, such as Oracle, still maintain a justified position in the market.

8 Likes
1489

I’d like to offer a bit of a layman’s view – and truly against better knowledge – that the winners mainly develop talent instead of trying to acquire as much of it as possible. And that the top experts in these winning companies would primarily handle internal development and training instead of working at the customer interface – while a disruptive transition is shaking the industry.

“Investment” at this stage would primarily be about how much effort is put into what could be called, perhaps old-fashionedly, “methodology development”: monitoring, researching, testing, acquiring, applying, refining, and then defining, implementing, instructing, training – as well as constantly expanding and updating.

And this would ideally happen on multiple levels, where at the core would be elements common to everyone, followed by function/industry/customer group-specific application work.

Of course, the more elite gurus a company has at its disposal, the faster and smarter it can transform. But perhaps even more essential is how much the company can and understands to invest in this.

If, on the other hand, a service company focuses on competing for “5,000 lines a day soloists” with the goal of selling them at the customer interface, then at least based on my layman’s gut feeling, things can easily go south. Or if they don’t go south, you’ll only see their taillights as these gurus move directly onto the customers’ own payrolls or subcontractor lists.

And certainly, the CTOs leading all this carry a heavy burden of responsibility. God help the companies where the CTO can’t keep up.

3 Likes
1490

Coding outsourcing may be under pressure because of AI. You just don’t need as many pairs of hands. You won’t pay much more for an AI coder if it’s a level playing field and everyone has the same AI tools, or if the AI tool provider takes their cut.

https://x.com/oguzerkan/status/2020025470784352727

1491

It is certainly interesting to see where this goes. A skilled individual is equivalent to a development team thanks to those tools, but where does this lead financially?

Skenaario 1: Roughly the same amount of work is done in total (there are many examples of this in technological development), fewer hands are needed, an oversupply of developers – a buyer’s market where the creators are the big losers – but it’s hard to say if the market is efficient enough for end customers to see savings?

Skenaario 2: More things get done, and in this case, consulting firms might continue with their current model; potentially, customers see savings (if more efficient work at the same hourly rate translates through to them).

Skenaario 3: Low/no-code solutions break through, and regardless of the volume of work, the share of developer consultants decreases, putting both the firms and the creators in trouble.

The AI hype perhaps leans more toward scenario two, but is it realistic in the long run? I can’t say. My guess is that simple tasks lead to scenario three, while the more difficult ones still require professional expertise, though they represent a shrinking piece of the pie.

Therefore, I would personally invest in some (likely new) AI-first product or consulting firms, rather than the current “body shops” (bench-warmer rental agencies).

3 Likes