The Great SaaS Rebundling: How AI Kills Point Solutions

The software industry moves in predictable waves. Understanding these patterns matters because we're at the start of another major shift—one that will reshape how companies build, buy, and use software.

Guillaume Cabane, founder of Hyper Growth Partners and advisor to companies like Ramp, recently mapped out how B2B software has evolved over the past two decades. His framework reveals a clear pattern: the industry swings between bundling and unbundling roughly every seven years, triggered by new technology and economic pressure.

Cloud Computing (2000-2008): Software Leaves the Box

Software stopped coming in boxes. Salesforce proved companies would trust their data in the cloud. HubSpot and others followed with single-purpose tools delivered over the internet.

The big shift: Distribution innovation, not product innovation.

Companies started bundling features because customers wanted fewer vendors. Why buy five tools when one platform could do it all? The first "all-in-one" platforms emerged.

Mobile & Social (2008-2015): Platforms Take Over

The 2008 crisis forced efficiency. The iPhone created new channels. This sparked the golden age of growth teams—Facebook, Uber, and Dropbox showed what data-driven marketing could achieve.

Platform bundling accelerated:

• HubSpot went from marketing tool to full CRM

• Zendesk added more customer service features

• The pitch: "Replace your messy stack with our platform"

Buyers liked it. Fewer integrations. One vendor. Unified data.

Data Analytics (2015-2022): Everything Breaks Apart

Then the pendulum swung hard the other way. Companies like Segment made it simple to move data between tools. This removed the main reason customers stayed locked into platforms.

Marketers could suddenly pick the best tool for each task. Need email? Twenty options. Want analytics? Dozens more. Best-of-breed became possible.

The martech explosion happened:

By 2024, over 11,000 marketing technology products existed. For context, only 150 existed in 2011.

Why unbundling took off:

• Data portability killed switching costs

• Specialized tools built better features

• VCs funded "best-in-category" solutions

But it went too far. By 2022, marketing leaders drowned in vendor relationships. Data existed everywhere, which meant it existed nowhere. Some companies had multiple CDPs competing to be their "single source of truth"—defeating the entire purpose.

The integration burden became impossible. Each tool needed procurement, implementation, training, maintenance. Attribution broke. Nobody understood the full picture.

AI Era (2022-Present): Bundling Returns

We're entering the fourth cycle. The pendulum swings back to bundling. AI provides both the technology and the business reason.

Why bundling makes sense again:

The complexity hit a breaking point. Managing dozens of tools isn't sustainable. The integration overhead exceeded the feature advantages.

AI changes the economics. Previously, platforms couldn't match specialized tools in every category. Now AI makes it dramatically easier to build sophisticated features quickly.

A platform can add capabilities that used to require dedicated companies. The feature gap between platforms and point solutions is shrinking fast.

What's happening right now:

• HubSpot and Salesforce add AI features directly

• New companies launch as comprehensive platforms from day one

• Acquisitions accelerate as platforms buy point solutions

• Marginal tools that don't provide 10x value will lose

Why Each Cycle Gets Faster

Each wave is shorter than the last:

• Cloud era: 8 years

• Mobile/social: 7 years

• Data analytics: 7 years

• AI era: probably 5 years

Three reasons for acceleration:

Technology adoption speeds up. Cloud took years to gain acceptance. Mobile went faster. AI achieved widespread use in months.

Execution gets faster. Building features used to require large teams and long timelines. Now AI lets small groups build sophisticated capabilities in weeks.

Market awareness spreads instantly. Competitors learn about successful approaches immediately through social media and conferences. The window to extract value before being copied keeps shrinking.

First-mover advantage matters more than ever. Competitive moats erode in months, not years.

Where We Stand Today

We're in early AI-driven rebundling. The shift is happening, but most companies haven't adjusted yet.

Look at the current martech landscape. Thousands of "AI-powered" point solutions are emerging—AI email writers, AI chatbots, AI analytics tools. This looks like innovation. It's actually the same unbundling pattern playing out again.

But this unbundling phase will be much shorter. A specialized AI email tool might have a 12-month window before major platforms ship comparable features natively.

The bundling is already starting:

• Large platforms aggressively add AI features

• HubSpot builds AI directly into their platform

• Salesforce does the same with Einstein AI

• New companies launch as full platforms immediately

The strategic shift: AI doesn't just enable better features. It enables owning the entire stack.

Because building software is dramatically easier with AI, platforms can expand horizontally much faster. Why let customers use ten tools when you can provide all ten capabilities in one integrated platform?

What This Means

The next five years will look more like 2008-2015 than 2015-2022:

• Consolidation instead of fragmentation

• Platforms instead of point solutions

• Integrated stacks instead of best-of-breed

For software buyers: The friction of managing dozens of tools will push you toward comprehensive platforms. Integration value will outweigh feature gaps.

For software builders: Unless your point solution provides dramatic advantages that platforms can't replicate with AI, you're building on borrowed time. The winners will offer comprehensive, integrated experiences powered by AI throughout.

We've been here before. The bundling wave of 2008-2015 followed the same pattern. The difference is AI accelerates everything. The cycle will complete faster. Winners will emerge sooner. Losers will have less time to adapt.

The rebundling has begun. The only question is whether you're positioned for consolidation or still playing the unbundling game from the last era.

Top Tweets of the day

1/

This video (on YouTube) gives a good understanding of why prompt engineering is over as AI can write prompts better than humans.

You just need money and a set of good examples + bad examples.

This is what the leading AI labs were talking about a year or 2 ago that the field of prompt engineering is over because the LLMs can write them much better. They probably saw what internal models could do.

Mike Taylor gives an example of good jokes versus bad jokes and it is easily one of the best joke creators I have seen in my life. You can literally create a stand-up out of it.

There will be new standup comedians, entirely based on AI prompting now. All you need is taste.

My mind is blown in just 30 minutes.

2/

Anthropic is showing a masterclass on how to build the new Apple.

It charges a premium price for AI and customers gladly pay for it even though 10x cheaper alternatives exist.

It is kinda like the Apple of AI.

3/

This is the new frontier. There are countless applications of multi-modal AI.

I once used it to write a post where the numbers were only mentioned in the video but the host delivering the talk didn't mention the numbers that were used in the slides.

It has various applications, like this one. You can even analyze MrBeast's videos and find the perfect analysis of what happens in his videos and train a custom AI on top of it to create a viral AI YouTube script writer. Many have created smaller ones like Subscribr AI which makes $100K MRR or something.

Rabbit Holes

1. Watch this for 9 minutes and you'll never get burnt out again

2. WTF is AI ?!

3. Let the LLM Write the Prompts

What’d ya think of today’s newsletter? Hit ‘reply’ and let me know.

Do me a favor and share it in your company's Slack #marketing channel.

First time? Subscribe.

Follow me on X.

More Startup Spells 🪄

1. Stealing from Adjacent Industry (LINK)

2. AI Video Creator's 30-Min YouTube Funnel Makes $42K MRR on Skool (LINK)

3. Turn Unstructured Text Into Insights With AI (LINK)

4. Astronomer CEO TikTok Scandal: How a Catfished Heiress Went Viral to Boost a Meditation App (LINK)

5. AI Caps as Pricing: How Anthropic's Claude Code and OpenAI's Codex Nudge Upgrades with Usage Limits (LINK)