The New “Google/Amazon” Will Be Built Differently. A Practical Guide for the new hunting season
The New “Google/Amazon” Will Be Built Differently. A Practical Guide for the new hunting season
Is it possible in 2025 to build a new Facebook or an even larger system of technology products? Yes, but the way of building has to change.
This article was born out of a comment thread discussing why it seems almost impossible today to build a new Google (here “Google” means any very large internet product).
There were many opinions. My view: it’s absolutely possible, but not in the way people usually imagine when they say, “Let’s build the next product X.” Especially now, when the Jobs To Be Done approach is more relevant than ever: to keep a customer satisfied, you need to solve several of their problems at once. Let me explain what I mean.
What do digital giants look like today?
We’re used to thinking that a unicorn (or decacorn) = one genius product. Facebook had Facebook. Uber had Uber. Airbnb had Airbnb.
But if you look at the list of the world’s largest tech companies, you’ll see the picture has changed. Most of them are no longer single-product businesses, but complexes: bundles of services, applications, and solutions that together cover a customer’s full set of needs or even an entire life sphere.
And here comes the tricky question: if that’s the goal, how do you actually build such a giant?
Play
Ecosystem ≠ One Product
Take Microsoft. They started with Windows and Office, but the market demanded more:
— Azure was created internally (in-house development)
— LinkedIn was acquired (M&A)
— GitHub was acquired (M&A)
— Activision Blizzard was acquired (M&A)
— Artificial intelligence and cloud solutions are being developed through partnerships (for example, with OpenAI).
It’s the same with other major players:
Amazon: built internally (AWS), acquired (Twitch, Whole Foods), expanded through partnerships (Prime Video). Google: built internally (Search, Ads), acquired (YouTube, Android), invested heavily through its corporate venture arm (Google Ventures), then partnered or acquired.
The result is not a single product, but a whole product line.
Corporate business as usual.
Can this logic become a strategy for the future? That’s where Complex Building was born.
I came up with the term for myself, to capture this concept — call it whatever you like :).
Complex Building is an approach where your (or your company’s) goal is achieved not with one product and not with one way of creating it, but with a combination of instruments.
The goal is defined in advance. It’s like playing LEGO: you know what you want to build, but each block is created differently.
And this, I believe, is how truly large and valuable companies should be built today. The size doesn’t matter: the approach works both for creating a new car brand and factory, or for something much narrower and niche.
It’s like the programmer’s axiom “you eat an elephant one bite at a time,” only applied to creating a major business, where multiple businesses can coexist within one “product.”
Example: Influencers in E-commerce
Suppose your goal is to cover the full lifecycle of an influencer in the school-age e-commerce niche (from beginner to superstar).
You want a regular person to:
learn how to become a niche influencer,
grow into a mid-level blogger,
become a superstar,
launch their own clothing brand for teenagers,
and throughout all these stages, remain inside your product ecosystem.
No one forces you to do this in a single day. The whole construction can take 10 years — the essence and approach don’t change.
To realize this plan, you’d need a whole set of services:
a school for beginners;
a monetization platform for micro-influencers in e-commerce;
an agency for the mid-tier segment;
payment and advertising tools for top influencers;
a boutique agency for superstars;
creation of your own clothing/accessory brands.
One product won’t solve this task. You need a complex.
And here’s where young entrepreneurs (and half of corporations) usually freeze: how do we build this? They call in their CPO/CTO, give them the task, and throw internal resources at it.
Play
The Full Toolbox: Beyond “Build or Buy”
Most companies get stuck in the binary debate of build vs buy.
In reality, the arsenal is much wider — here’s a complete list (if I missed anything, add in the comments):
1. Internal formats (full control, resources inside the corporation)
— In-house development — internal engineers and designers building new products. Maximum control, but limited speed.
— Internal department (shared resources) — current staff taking on new product tasks in addition to their main role. Cheap, but focus suffers
— Internal studio / innovation lab (“garage”) — small unit that turns ideas into prototypes and MVPs
— Spin-off — creating a separate legal entity to give the team more freedom or prepare for external investment.
2. External market formats (working with independent companies)
— Incubators — programs that turn ideas into first product versions. Usually external
— Accelerators — short-term programs to test startups against corporate strategy. — Contests and hackathons — fast way to attract teams and talent
— Crowdsourcing platforms (e.g., Kaggle, InnoCentive) — global calls for solutions
— Open Innovation / Venture Client — the corporation becomes a customer of startups, testing their solutions without investing
— Product licensing / white-label — reusing or rebranding an existing product.
3. Investment formats (capital, not always control)
— Corporate venture fund (CVC) — investing in startup equity without operating control. Access to innovation markets
— LP (Limited Partner) in an external VC fund — invest in a VC fund to gain access to its startup pipeline
— Joint venture — a new legal entity co-founded with another company to share risks and rewards
— Mergers & Acquisitions (M&A) — buying and integrating a company or product. Fastest but most expensive route.
— Venture Studio as a Service — an external team that builds startups “turnkey” for corporations
— Corporate Accelerator-as-a-Service — an external team that runs corporate accelerator programs end-to-end
— Corporate Innovation Scouts — “innovation scouts” systematically searching for technologies and startups globally
— Strategic partnerships — long-term collaborations without equity ownership, but with access to partners’ resources/markets.
5. Future and ecosystem formats
— AI agents — still experimental, but likely to become tools for idea generation, hypothesis testing, and even partial autonomous launches
— Ecosystem platforms (ecosystem play) — building platforms (like Android or AWS Marketplace) that attract startups and services
— Combinations of formats — in practice, it’s almost always a hybrid: projects enter through partnerships, accelerators, or funds, and exit via M&A or spin-off.
The magic of Complex Building (and the demand of our time) is that you use all these simultaneously, choosing the most fitting method for each part of the puzzle. And each part can — and should — be an independent business.
Applying it to the Influencer Example
School for influencers — easier to launch through a startup studio (venture builder)
Monetization platform for beginners — through in-house development
Agency for mid-tier influencers — as a joint venture
Advertising and payment tools for top influencers — via M&A
Boutique agency for superstars — through strategic partnerships.
Any block can fail. In that case, you either iterate or change the creation method. But the overall goal doesn’t change: you need all the “blocks,” ideally built in parallel and eventually connected into one large player.
And even if some blocks fail, you’ll still have one or two working independent businesses — much better than one giant half-built spaceship.
Benefits of the Approach
— Speed. Don’t waste years developing what you can buy. Launch pieces of the business right away
— Flexibility. Risky ideas are tested through a startup studio, not the entire corporate P&L
— Cost efficiency. Simple things are faster and cheaper to build internally
— Control. Keep key levers — and, most importantly, the knowledge of how the blocks fit together
— Focus. Complex Building turns a chaotic product portfolio into a manageable strategy
— Monetization. Each module can and should be its own business. Easier to fix: if something fails, replace it with a cheaper alternative (e.g., failed monetization → switch to a partner’s tool)
— Stealth. Competitors don’t immediately see how they’re being surrounded
— Expertise. High-risk, high-innovation elements are more realistic to build via startups and partnerships than inside a corporation.
Core Thesis
If the task is, for example: “Our business is a portfolio of products that covers the entire lifecycle of an influencer in the e-commerce niche for school students (from beginner to superstar).”
Then instead of piling features onto a single product and trying to build a “spaceship,” Complex Building raises your odds of success — and of creating something truly innovative.
For such a massive undertaking, one or two good ideas are not enough, and a single team can’t deliver everything. You need to gather startups from the niche (each with one good idea), and together something substantial may emerge.
Play
Pitfalls
Of course, the approach has its downsides:
— Management stretch. One product is easy; managing five creation formats is harder. Few people know them all well
— Cultural conflicts. In-house teams and joint ventures think differently
— Integration. Acquisitions take months to merge with internal processes
— Finance. Complex Building means constant trade-offs — what’s cheaper, faster, or more strategic. Mistakes are costly, but still cheaper than trying everything in-house and writing it all off.
In almost every country, there are stories of attempts to build “national” search engines, marketplaces, social networks, or messengers in a top-down way. Most didn’t work. The ambition was huge, but building everything from scratch with one team in a green field is… let’s just say, a brave task.
Conclusion
The big companies of the future will increasingly not grow out of a single app. They will be assembled like puzzles: some parts built in-house, some acquired, some licensed, some developed via partnerships — but all unified by a clear design and strategy.
Complex Building also makes companies less vulnerable to failure: if one block fails, you still have the others (since each block is itself a business).
This is the alternative to today’s pattern: one product grows → “what can we bolt on?” → feature factory.
Complex Building is a language and a method for deliberately assembling such complexes.
It’s not about vague “ecosystems” (a nice word with little substance), but about a specific set of choices: this part we build (in-house), this we buy (M&A), this we invest in (Corporate VC, joint venture), this we partner on (strategic partnerships, licensing, open innovation).
And the acceptance that one customer value is not enough anymore — you need to cover a wider set of customer jobs (Jobs To Be Done).
Doing all this in-house was hard before; now it’s harder still.
With the AI era, it first seemed like everything would get simpler: give a task to an agent and relax. But in reality, the complexity has only shifted: from “write tons of code (slow and expensive)” to “figure out what exactly to build, how the parts connect, and why.”