Toward the end of last year, a friend who was thinking about leaving an in-house role for an IT services or consulting firm asked me a question:
“Everyone keeps saying Japanese companies are past their prime. Most seem to be barely growing their sales, if they’re growing at all. So why have the companies supporting them—IT vendors and consultancies—been growing so much over the past few years?”
I remember hearing much the same question from someone else who had made that move.
When you work at a Japanese systems integrator, or “SIer,” consecutive years of record revenue and profit start to feel normal. A year without revenue growth would be a serious cause for concern. Apparently, that looks rather enviable from other industries.
For readers unfamiliar with the term, SIers are companies that build, integrate, and often operate IT systems for other businesses. In Japan, they handle a substantial share of the work that companies elsewhere might give to their own technology departments.
Here is what I told my friend:
“Japan’s economy is like an old man who has retired and been hired back. He’s being given far more IV nutrition than he needs every day to keep him going. But he doesn’t actually have much use for it anymore. So the IT vendors and consultants living inside his body start feeding on it and multiplying like cancer cells.”
A little harsh, perhaps. But I think it captures the situation rather well. Yes, I’m pleased with my own analogy.
So why can these firms grow so quickly in an economy that seems to be going nowhere? Here is how I see the mechanism.
After Three Decades of Stagnation, the Economy Enters Old Age
Everything that rises eventually falls. I think national economies have life cycles, too.
For the purposes of this argument, let’s take that as given. The medieval storytellers of The Tale of the Heike were saying much the same thing centuries ago, and I’m happy to borrow their authority. Looking at other countries, it doesn’t seem an entirely unreasonable premise.
In my reading, Japan reached its peak somewhat later than some other developed economies, then spent decades stagnating. Those are the “lost decades” you keep hearing about. Now some of the countries that grew early among the developing economies seem to be reaching their own turning points.
If the Japanese economy were a stereotypical office worker, it would be the older guy counting down to retirement in a quiet corner of the department. Or the retiree brought back on a reduced contract who comes to the office every day without doing very much.
Left alone, demand and supply would keep shrinking. Businesses and households would have less to spend. Decline would feed on itself.
One response was Abenomics, the economic program associated with former prime minister Shinzo Abe. Its monetary-policy component pushed interest rates down and flooded the financial system with money, with the aim of getting that money circulating through the economy.
Back to our old man. He spent his entire life working and never developed any hobbies. Once he stopped working at full speed, his health deteriorated. So now he gets a daily IV drip to keep him out of bed and looking reasonably lively.
In the limited sense that companies gained access to readily available funding, Abenomics worked. Lower interest rates also helped weaken the yen, boosting profits for exporters, including manufacturers.
Companies had more money. What they often lacked was somewhere attractive to invest it in their core businesses.
Their markets were already mature. They weren’t necessarily about to develop a breakthrough technology, build another store or factory, or enter some promising new domestic industry. Nor were they eager to raise their labor costs by investing more in people.
So they tend to arrive at a few familiar options:
- Launch a new business or buy an overseas company, like buying a lottery ticket—even when you can practically see the disappointment coming.
- Invest in IT to improve efficiency and margins, because growing revenue in the core business looks difficult. The AI boom is the most obvious example.
- Commission a grand, uplifting management initiative, packaged as a beautiful PowerPoint presentation, because the actual future looks too bleak to discuss. Corporate fairy tales, essentially.
And then, seemingly out of nowhere, the IT vendors and consultants arrive. They find all sorts of ways to collect that money. The client has run out of ideas, so it ends up going along with what it is sold.
Repeat that across Japan’s millions of companies, and it becomes easier to understand why the firms selling them help are doing so well.
To stretch the old-man analogy a little further: the treatment has restored some of his energy. He has money and time. But by this point, he no longer feels he has a place at work or at home, and there is nothing in particular he wants to do all day.
The nutrients from the IV keep circulating without much purpose. The growths that had been sitting quietly inside him wake up, feed on the surplus, and start multiplying.
Even the Healthy Cells Want In
Recently, I heard a company announce in a television commercial that it was, in effect, now an SIer.
That surprised me.
When I was looking for my first job in the late 2010s, Japan’s layers of IT subcontracting were regularly blamed for weakening the country’s competitiveness. SIers were treated as something to be despised—or, at best, a necessary evil.
So hearing a company proudly use the term in a national TV commercial was quite something.
My other reaction was: “You lot are getting into this business too?”
The friend I mentioned earlier works at a consumer-facing technology company. Apparently, that company is expanding its business-to-business operations. I have started hearing its name occasionally as a competitor.
It isn’t the only one. Plenty of companies seem interested in moving from running their own products and services to helping other companies run theirs.
Of course, the distinction is fuzzy. Dividing the world into “operating companies” and “supporting companies” is probably something only people in the latter group bother doing. Most businesses have elements of both. Perhaps the categories themselves aren’t especially useful.
Still, within my admittedly unpleasant analogy, more formerly healthy cells seem to be deciding there is money to be made as a tumor in the Japanese economy.
And why wouldn’t they? It pays.
There Is Something Odd About the Helpers Outgrowing the Businesses They Help
I don’t think this is easy to classify as simply good or bad. But as someone working on the supplier side, I can’t help feeling uneasy about a system in which the companies providing support keep prospering.
It does keep me employed. I’m aware of that.
Even so, let me be blunt: if management has run out of ideas and is outsourcing strategy or IT simply because it cannot think of anything else to do with the money, it will not work.
AI in particular requires a clearly defined problem, a strategy, suitable data, and sustained effort. Starting a project because generative AI happens to be fashionable is a good way to pour money down the drain.
Perhaps my own surroundings are unusually disappointing. Perhaps AI is delivering substantial business value elsewhere in the world, and the investment case is much stronger than it looks from where I sit.
But there is one feature of the current spending boom that, to me, makes the bubble hard to deny.
I work on business AI implementations every day, and measuring their benefits quantitatively is exceptionally difficult—even compared with other IT initiatives. I don’t see why that difficulty would be unique to Japan. The spread of generative AI has made calculating the return harder still.
Yet executives keep committing money, year after year, on a scale that would cause serious trouble if it failed to pay back. They cannot clearly quantify the expected return, and often cannot tell whether they have earned it after the fact either.
I find it hard to call that normal.
Money with nowhere better to go ends up flowing into projects where nobody knows whether it can be recovered—or even whether it already has been.
Perhaps AI’s vague, qualitative benefits are rather convenient for executives who have exhausted their options. If the payoff cannot be pinned down in monetary terms, there is less pressure to explain exactly what the money achieved.
My suspicion is that a great deal of what we are seeing works like this: AI executives, investors, and evangelists sell an abstract future with very little substance, encouraging companies to put their surplus cash into AI and helping push AI-related valuations higher. Mature businesses with no obvious next move buy the story and hand over the money.
So what should happen instead?
In principle, companies that have reached the end of their useful life should be allowed to wind down sooner, freeing people and capital for new businesses.
But that would mean considerable pain, including large numbers of people losing their jobs, even if only temporarily.
I also happen to think humanity itself may have passed its peak and entered a period of decline. From that perspective, I’m not convinced there is an endless supply of promising new investments waiting to absorb those resources. Break up the old companies, release their people and money, and they may simply be drawn into the next illusion, and then the one after that.
Which leaves us prolonging the lives of what I would call zombie companies: businesses kept going by abundant money and, particularly among exporters, the weaker yen that low interest rates helped produce.
As I write this, Japan is moving toward higher interest rates and attempting to correct the yen’s weakness. But when these companies support so much employment, policy cannot squeeze them hard enough to kill them off in large numbers.
That leaves a role for IT vendors and consultants. They take on central functions such as management support and technology, help keep the underlying businesses alive, and collect the surplus money along the way.
A necessary evil, perhaps.
That is very much the supplier’s argument, of course.
What We Need First Is to Admit Where We Are in the Life Cycle
Does that mean I think it is fine for this arrangement to continue, with suppliers drawing money out of client companies that have run out of options?
No.
While companies keep paying for fantasies and theories that exist only on presentation slides, the parts of their businesses that still have value can deteriorate too. Follow this structure far enough, and the benign tumors could turn malignant, eventually consuming their hosts.
I can only speak with any confidence about IT. But even within data and AI, there are investments that resemble lottery tickets and investments that resemble a sensible, diversified fund. I think that distinction applies to IT spending more broadly.
A company in its fifties or sixties should be thinking about prudent asset management for its later years, instead of buying lottery tickets and dreaming of a jackpot. There is such a thing as an IT strategy appropriate to that stage of a company’s life.
Some of the spending I see at long-established Japanese corporations reminds me of struggling rural municipalities near the end of Japan’s asset bubble. They knew their future tax base was disappearing, but used the money they still had to build accommodation and tourist facilities nobody needed.
The same problem applies to how companies deal with their outside advisers and IT suppliers.
If rapid growth is no longer a realistic prospect, the guiding question should be how to keep the company stable and viable for as long as possible with the people it already employs. That should shape what it asks its suppliers to do.
The suppliers have responsibilities here too. Flashy initiatives certainly move more money. But if clients eventually lose the ability to earn a living, the firms serving them will lose theirs as well.
We should be helping customers over the long term, without letting fashionable technologies or short-term economic movements dictate everything.
A company can still be healthy at seventy or eighty. I would like to see clients and suppliers build realistic relationships that suit each stage of that life, allowing businesses to reach a respectable old age.
For 2026, I want to think beyond the next sales figure or the success of the next project and take a longer view myself.
That all sounds very constructive. But, honestly, what I am feeling right now is mostly anger at executives and investors who go around declaring how well their AI investments are working.
People appear on television, string together some impressive-sounding words about AI that mean absolutely nothing, and convince business leaders to believe them. Then engineers like me, working at the firms expected to implement the thing, have to deal with the consequences.
Having work is good. I know.
But the gap between what people expect from AI and what it can actually deliver is getting out of hand. Thinking about the projects waiting for me after the New Year is already making my stomach hurt.
Until next time.
This article is an English translation and adaptation of “SIerやコンサルは日本経済の良性腫瘍！？支援会社の売上が増加する仕組み,” originally published in Japanese on pickerlab.net on January 3, 2026.
Comments