AI productivity tools: How the second-brain market is taking shape

Editor’s Note

This article was adapted by EUVC from an original insight published by Synaptic.

Read the original article.

The idea of a “second brain”, popularised by Tiago Forte, describes an external system for organising and retrieving knowledge. Generative AI gives that concept a more literal interpretation: software that can understand context, surface relevant information and, in some cases, act on a user’s behalf.

Synaptic’s market analysis argues that this shift is creating a broad category of AI productivity tools. These products are not tied to a single business function. Instead, they support everyday work across knowledge management, communication, planning and workflow execution.

The market spans large technology platforms, established productivity companies and more than 100 specialised tools. The central question is where durable value will accrue as foundational AI features become standard.

AI productivity tools are function-agnostic products designed to improve routine work, including:

Their defining feature is context. Rather than responding only to a generic prompt, these tools can draw on a user’s notes, conversations, documents or activity.

This makes generative AI particularly relevant to knowledge work. McKinsey’s analysis of generative AI highlighted knowledge management as an important organisation-wide opportunity, alongside applications in customer operations, marketing, sales, software engineering and research and development.

Synaptic divides the emerging product landscape into two broad interfaces.

AI assistants answer questions, retrieve information, summarise material or create content using a user’s prompt and personal context.

The user remains responsible for deciding what to do with the output. Examples in Synaptic’s analysis included chatbots such as ChatGPT, knowledge assistants such as Mem and AI search engines such as You.com.

AI agents act directly through integrations with calendars, email clients and other software. They typically break an instruction into several steps and make a series of calls to large language models and external systems.

Synaptic’s examples included task-automation products, scheduling agents and email agents.

At the time of the analysis, assistants were more established because they preserved a clear layer of human oversight. Agents offered a more ambitious proposition but remained relatively expensive and insufficiently reliable for unsupervised work.

Synaptic identified Microsoft and Google as the defining competitors in the AI-enabled workspace.

Microsoft integrated Copilot across Windows, Microsoft 365 and its browsers. Google responded by adding generative AI capabilities to Workspace, then branded as Duet AI. At publication, Synaptic also anticipated Apple’s entry into the category.

The significance of this competition extends beyond the platforms themselves. Features embedded by major technology companies can quickly become standard expectations rather than sources of differentiation.

This shapes the opportunity for startups. As the large platforms establish the core layer, new companies must build higher-order functionality around particular workflows, users or industries.

That tension also informs the wider debate about how AI may change SaaS rather than simply replace it.

Synaptic identified three main dimensions of specialisation.

Vertical products adapt AI to the terminology, workflows and accuracy requirements of a specific sector. Examples included DeepScribe in healthcare and RemNote in education.

The potential advantage is domain relevance: a focused product can be designed around needs that a general-purpose assistant may not address.

Other companies concentrate on a single recurring task. Examples included Otter for meeting transcription, xTiles for brainstorming and Reclaim for calendar scheduling.

These products compete through workflow depth rather than breadth.

Products also differ in the kinds of information they process. Some focus primarily on text and written knowledge, while others combine audio, video and text.

For example, Synaptic contrasted text-oriented products such as Mem with meeting tools such as Fireflies, which connect recorded conversations with written summaries and workplace systems.

Synaptic argues that AI productivity software offers a particularly credible route from individual use to enterprise adoption.

The lower end of the market contains a wide range of individual and small-business needs. This gives startups an environment in which to attract early users, test their products and refine their positioning before pursuing larger organisations.

Productivity software can also spread from the bottom up. Individual employees adopt a useful product and subsequently introduce it to their teams or employers. Notion and Airtable are examples of companies that followed this pattern.

Enterprise adoption matters because integration into organisational workflows can make a product harder to replace. However, moving from individual utility to company-wide deployment is not automatic. The related EUVC article on the GenAI divide between individual value and enterprise-scale adoption examines this challenge in more depth.

Synaptic described an active funding environment across stages, from pre-seed to Series F.

In its dataset at the time:

Synaptic linked this concentration of capital to the potential durability of enterprise software. Once a tool is embedded in workflows and connected to organisational systems, it can become increasingly important to day-to-day operations.

These figures should be read as a snapshot of Synaptic’s dataset rather than a current estimate of the market.

Synaptic’s website-traffic analysis captured both initial enthusiasm and an early decline in attention.

Approximately 72% of the startups in its dataset recorded positive quarter-on-quarter website-visit growth in Q1 2023. Of those companies, 75% experienced a decline in Q2.

Synaptic interpreted this as a possible “trough of disillusionment”: the period in which early excitement gives way to a more practical assessment of implementation, adoption and value. ChatGPT followed a similar pattern, with Similarweb reporting an approximately 10% decline in global traffic in two successive months during Q2 2023.

Several note-taking and task-management products nevertheless recorded strong growth:

Synaptic’s broader expectation was that engagement would become less driven by novelty and more dependent on demonstrable value.

Synaptic highlighted Process Street, Superhuman and Spike as product leaders in its analysis.

Process Street applied generative AI to process management, including workflow creation, task instructions, approvals and integrations with other workplace tools.

Superhuman incorporated AI into email composition and editing, including drafting messages in the user’s voice and supporting research inside the email workflow.

Spike combined a chat-like email interface with AI-generated replies, email summaries and an assistant embedded within the inbox.

The common thread is workflow proximity. Each product placed AI inside an activity users already performed, rather than requiring them to move into a separate general-purpose interface.

Synaptic’s analysis points to four enduring questions.

The competition between major workspace platforms will determine which capabilities become baseline features. Startups need to build beyond functionality that Microsoft, Google or other platform providers can distribute natively.

Low barriers to entry and a large number of products make product-market fit difficult. Domain expertise, workflow depth, proprietary context and strong integrations may help companies distinguish themselves, but Synaptic did not present any of these as a guaranteed defence.

Synaptic identified Superhuman and Mem as products beginning to demonstrate enterprise adoption. It also highlighted Fireflies, Spike and Taskade as companies to watch as the funding market developed.

These assessments reflect the market at the time of Synaptic’s analysis and should not be treated as current investment recommendations.

A second brain requires access to significant amounts of personal or organisational information. Adoption may therefore depend on transparency about how data is used, alongside appropriate regulatory frameworks.

Accuracy is equally important. Hallucinated answers, outdated information and errors in task execution make users cautious about delegating work without oversight.

It describes an external system that can organise and retrieve a user’s knowledge. Generative AI extends the concept by interpreting context, generating responses and potentially taking action.

No. Assistants provide information or content for a user to act on. Agents connect to other systems and execute tasks directly, often through a multistep process.

Synaptic’s analysis suggests focusing on specialised verticals, deeper use cases, different modalities or workflows that sit beyond the core capabilities offered by large platforms.

The central barriers are reliability, cost, data privacy, transparency and the risks of allowing AI systems to act without adequate human oversight.

Synaptic is a private-market data platform that turns company information, including financials, decks, websites, news and alternative metrics, into connected data for investors.

Learn more about Synaptic’s work supporting data-driven private-market research on their website.

This article was adapted by EUVC from an original article published by Synaptic.

Read the original article.