1 Billion Users for Gemini – and the License Often Still Goes Unused

On August 11, 2026, Sundar Pichai announced a figure that's unusual even by Google's standards: the standalone Gemini app crossed the mark of one billion monthly active users. In his announcement on X, Pichai called Gemini the fastest-growing product in the company's history – and the 14th Google service ever to break the billion-user threshold. For a company that already runs several global mass-market products in Search, Gmail, Maps and Android, that's a remarkable milestone.

As a business owner or decision-maker in the Mittelstand, this news may at first seem like exactly that: news from Silicon Valley, far removed from your day-to-day operations. But a second look is worthwhile – not because of the growth figure itself, but because of what sits right next to it: your company is, in all likelihood, already paying for exactly this product without realizing it. And the available data shows that most companies barely use it anyway. That's exactly where this article picks up.

The instinct to react to a headline like this is understandable: when a product grows from 400 million to over a billion users within a few months, it's easy to feel you need to act now to avoid falling behind. But in many companies, that exact instinct leads to an expensive mistake: buying yet another AI tool before even clarifying what's already available in the tools you already have. This article therefore puts the numbers in context, separates the genuinely solid figures from the promotional ones – and derives the two questions you should ask yourself before any further AI investment.

The moment: how Gemini became Google's fastest-growing product

The growth curve behind this milestone is unusually steep. In May 2025, the Gemini app had roughly 400 million monthly active users. By around May 2026, that figure had more than doubled to 900 million. At the Q2 earnings call on July 22, 2026, Google cited 950 million monthly active users. And less than three weeks later, on August 11, 2026, the billion mark had been crossed.

Those final 50 to 100 million users in under three weeks are the real reason Pichai calls this the fastest growth story in the company's history. Google itself attributes part of it to changing usage patterns: 63 percent of Gemini users now use voice control, and in Gemini Live, one in five sessions now includes a camera or screen share rather than pure text input. Gemini is increasingly being used as a multimodal assistant – not just a chat window.

The context around this figure is notable. ChatGPT had already crossed the one-billion mark back in June 2026 – though measured as weekly active users, while Google reports its figure as monthly active users. Several sources explicitly point out that these two metrics aren't methodologically directly comparable. Drawing a "tie" between the two providers from this oversimplifies the picture. For your own business assessment, it matters less who is "winning" anyway than what these numbers say about the actual market penetration of AI assistants.

That market penetration is real, and it's still accelerating. In just over a year, the number of people regularly interacting with an AI assistant like Gemini or ChatGPT has more than doubled. For personal use – research, drafting text, quick answers to everyday questions – this has become so commonplace by now that it barely needs explaining. For business use, that's explicitly not the case. And that's exactly where the confusion begins that underlies many decisions in the Mittelstand.

The risk of confusion: an app number is not a business number

Before we get to the actually relevant question, one clarification is needed that gets lost in many short news items on this topic: the billion figure refers exclusively to the standalone Gemini app – the app users actively download and open, similar to ChatGPT. It explicitly does not include the "AI Mode" in Google Search, for which Google separately reports over a billion monthly active users worldwide. It also doesn't include Gemini usage within Google Workspace – that is, in Gmail, Docs, Sheets, Slides or Meet. And it says nothing about Gemini's integration into Android devices.

This distinction is crucial for your own assessment. If someone at your company says "a billion people use Gemini, we need to use it too," that statement refers to a consumer product, not to the question of whether Gemini actually does anything for your daily business – in your accounting, your customer service, your quoting process. For that question, a completely different set of numbers is far more relevant.

Google itself has not yet published a separate, dedicated user figure for "Gemini in Google Workspace." Individual industry blogs circulate estimates of 15 to 20 million paying enterprise seats by the end of 2026 – but that figure is a projection from individual analysts, not an official Google statement, and should be flagged as an estimate when discussed with colleagues or advisors, not treated as a proven number. Anyone relying solely on publicly available figures to assess Gemini's success in a business context is on thin ice. What can be documented, however, is the pricing structure through which Gemini has become embedded in practically every larger Google Workspace contract today – and that's exactly worth a closer look.

What many SMEs overlook: you're already paying for Gemini access

Since early 2025, Gemini's AI features have been built into the Business and Enterprise plans of Google Workspace – in Business Standard, Business Plus, and the Enterprise Starter, Standard and Plus tiers. Google discontinued the previously separate add-on packages "Gemini Business" (20 US dollars per user per month) and "Gemini Enterprise" (30 US dollars per user per month) and folded the cost into the base prices instead. In turn, base rates rose by 2 to 4 US dollars per user per month.

In practice, that means: anyone using Google Workspace Business Standard in 2026 – currently around 14 US dollars per user per month on an annual plan – already has full access to Gemini in Gmail, Docs, Sheets, Slides and Meet. No separate upgrade, no additional approval process, no separate invoice. For many small and medium-sized companies in Germany that already use Google Workspace as their standard office software, that means: AI access has long since been paid for. The only question is whether it's actually being used.

From Google's perspective, this shift is a logical one: instead of selling an add-on product that has to be separately introduced, communicated and budgeted for, the AI feature becomes part of the base package – much like spell-check or cloud storage once were. But for your company, it also means: the decision "should we adopt AI?" has effectively already been made for you, without anyone at your company actively saying "yes." That's exactly why it's worth taking a quick inventory of your current Google Workspace plan: if you use Business Standard, Business Plus or one of the Enterprise tiers, you're already paying for Gemini access through the base prices, which rose by 2 to 4 US dollars per user per month – regardless of whether a single employee has ever even opened the feature.

The central finding: a paid license is not a used license

And this is exactly where the core of this article lies. A survey cited by Recon Analytics of more than 150,000 enterprise users from 2026 measured what's called the "Workplace Conversion Rate" – the share of employees who have access to an AI tool but actually use it actively. The result for Gemini: 34.0 percent. For comparison, Microsoft Copilot came in at 35.8 percent – a similarly low figure. ChatGPT, on the other hand, reached 83.1 percent.

Only about a third of employees with paid Gemini access actually use the tool actively – two out of three paid licenses effectively go unused.

This figure should trigger a specific question at your company: how many of your employees with a Google Workspace Business or Enterprise license have even opened Gemini in the last 30 days? The Recon Analytics study comes from a smaller analyst firm whose methodology hasn't been independently verified – it should therefore be cited with appropriate caution, not treated as absolute truth. As a prompt for an internal stock-take, though, it's valuable, because it confirms an observation many business owners already know from experience: features built into software are often simply overlooked if nobody actively demonstrates what they're good for.

It's also notable that Gemini's 34.0 percent conversion rate isn't an isolated case. Microsoft Copilot, similarly deeply integrated into Office 365 and following a comparable "bundled rather than actively introduced" approach, sits at nearly the same level at 35.8 percent. That suggests the low usage rate isn't a Gemini-specific problem, but a structural pattern among AI features deeply embedded in existing software: they ship by default but are rarely actively rolled out, trained on, or built into concrete workflows. ChatGPT, on the other hand, which most users have consciously chosen as a standalone tool and integrated into their daily routine, reaches a usage rate more than twice as high at 83.1 percent. The difference, then, has less to do with the technical quality of the respective models than with how the tools were introduced into everyday work.

Why employees choose ChatGPT anyway when given a free choice

The Recon Analytics survey becomes even more revealing when you look at what happens once employees have access to Copilot, ChatGPT and Gemini at the same time and have to choose a primary tool. The result: only 8 percent choose Copilot, 18 percent choose Gemini – and 70 percent choose ChatGPT.

This distribution is more than a footnote. It suggests that an AI assistant embedded in the office suite – whether from Google or Microsoft – isn't the first choice for many employees once a real alternative is available. One obvious reason: Gemini in Gmail or Docs is designed as a reactive, general-purpose chat assistant. It answers questions, summarizes text, drafts emails. What it typically isn't: an agent tied to a specific, recurring business process – reviewing incoming invoices, generating quotes according to a fixed template, or structured customer communication along a defined workflow, for example.

That exact gap – between a general chat assistant built into the suite and a structured process agent with access to your actual company data and workflows – is the point where the real business decision arises for the Mittelstand. Not "which AI tool do the most people worldwide use?" but "which recurring task in my business can't be solved by a generic chat in the inbox, but needs an agent that independently handles an entire workflow?"

An example makes the difference tangible. Gemini in Gmail can help you phrase a single customer email more politely or summarize a long reply. What it won't do on its own is check newly arrived invoices every morning, reconcile them against your orders, and forward only the exceptions to you. It won't independently generate a quote based on your pricing logic the moment an inquiry comes in through your contact form. And it won't prepare next Monday's schedule on its own every Sunday evening. All of that requires an agent tied to a specific workflow, with permission to access your data, that acts without a manual trigger. This distinction – reactive assistant versus proactive process agent – explains why the billion-user news from the consumer space carries little direct weight for your operational reality.

What you should take away from this figure for your company

Google additionally points to figures showing that enterprise users save an average of about 105 minutes per week through Gemini tools, and that 75 percent of daily Gemini-for-Workspace users report improved work quality. The claim that Google customers edited more than 1.4 billion Google Docs files using Gemini's generative AI feature in the first half of 2025 also comes directly from Google. These figures aren't wrong, but they come from Google-adjacent or marketing sources and haven't been independently verified. In your own decision-making, they should be treated as an indication, not as proof.

What can be reliably established, on the other hand: first, Gemini has achieved enormous reach as a consumer product, independent of business usage. Second, Gemini is already included and paid for in many Google Workspace plans that German SMEs already use. Third, the available enterprise data shows that a paid license often doesn't translate into active use in practice – true for Gemini just as much as for Microsoft Copilot. And fourth, once employees have a choice, the majority pick a standalone, dedicated tool over the built-in assistant.

One further point of context is also important, one that's often missing from coverage of this topic: there are, as yet, no concrete, independently verified figures on actual Gemini usage specifically among German or European small and medium-sized companies. The cited adoption figures come from global, predominantly US-heavy enterprise surveys and don't translate one-to-one to a German craft business, a tax firm, or a mid-sized machine-building company. That doesn't argue against the underlying observation – built-in AI features are often used less than expected – but it does argue against uncritically applying the exact percentages to your own company, and for determining the number that actually matters for you instead: how many of your employees actively use the AI features you're already paying for?

Two very concrete next steps follow for you, and you can carry both out yourself without much effort. First: check internally how many of your employees with a Google Workspace Business or Enterprise license have even opened Gemini in recent weeks. This information can usually be found in your Workspace admin console. Second: list the recurring tasks in your company that are still handled manually today, even though they follow a clear, repeatable pattern – invoice review, quote generation, appointment coordination, initial contact with prospects. For these tasks, the decisive question isn't which chat assistant is built into your software, but whether a structured agent can independently take over the entire workflow.

This is exactly where sensible consulting comes in: not selling you another AI subscription, but working with you to determine what's already sitting unused in your existing licenses – and where an agent tailored to your business would actually make a measurable difference. The billion-user news out of Silicon Valley isn't a reason to rush, but a good occasion for a sober stock-take.