Lindy Review 2026

Nobody can keep up with AI productivity software anymore. A new tool ships every week, each one promising to automate your work, capture your tasks, and run your day for you. But shipping fast tells you nothing about whether a tool will still be around next year. The Lindy Effect is a handy way to tell the tools built on lasting workflows apart from the ones riding a temporary wave of attention. This review applies that thinking to AI productivity in 2026, and explains why this+that’s inbox-first approach lines up with the patterns of work that tend to stick.

Key Takeaways

• The Lindy Effect is a filter for picking AI tools in 2026. With non-perishable technologies, each extra year a thing survives raises how long you should expect it to keep going. That gives you a way to tell lasting productivity software from short-lived hype.
• AI sharpens the Lindy test instead of breaking it. Once building software gets easy, the fact that something has survived counts for more, not less. A company that lasts 8-10+ years has a structural toughness AI cannot conjure up overnight.
• this+that is built on Lindy-compliant foundations. Email (SMTP, 1982), the basic ideas behind task management, and integration-first architecture each carry decades of proven utility, and this+that adds AI on top of them rather than throwing them out.
• Much of a business’s failure risk concentrates in its early years. A tool or platform that makes it past that window has earned a kind of market validation that funding and hype can’t fake.
• The inbox is the most Lindy-compliant productivity interface there is. New apps pile up every day, but email and messaging have outlived cycle after cycle of technology. An inbox-first approach matches where work already happens.

Most of the AI productivity tools that launched in 2023-2024 will be gone by 2028. That isn’t doom-saying; it’s just what the pattern predicts. The Lindy Effect, which has guided technology and investment decisions for decades, says that the longer a non-perishable thing has already lasted, the longer it’s likely to keep lasting. Keep that in mind while you evaluate AI task capture platforms in 2026 and the durable choices start to separate from the expensive experiments.

this+that takes a different route. Instead of being one more wrapper around a large language model, it runs inbox-driven task execution across the tools you’ve already connected. The real question is whether that approach lines up with the Lindy-compliant principles that tend to predict long-term survival.

this+that: AI Personal Assistant for 2026

Grand View Research put the productivity management software market at USD 59.88 billion in 2023, with a projection of USD 149.74 billion by 2030. And yet the average knowledge worker still bounces between 11+ applications a day. The integration-platform market keeps climbing too, with iPaaS revenue estimated to have risen from $5.9 billion in 2022 to more than $9 billion in 2024. The contradiction points at something basic: more tools haven’t added up to less work.

The Lindy Effect helps explain why. Nassim Nicholas Taleb gave the idea its formal shape: for non-perishable things like technologies, ideas, and business models, how much longer you can expect them to live is proportional to how old they already are. A book that’s been read for 100 years will probably get another 100, and a programming language that’s been in use for 50 will probably stick around for 50 more.

What makes something Lindy-compliant:

• Non-perishable nature: It’s an idea, a technology, or a system, not a physical object that wears out
• Survival through stress: It has been exposed to market pressure, competition, and shifting conditions and kept going
• Revealed preference: People keep using it, which is the clearest sign it still delivers value
• Low hazard rate: Its odds of failing drop with every year it survives

The 2026 relaunch of this+that puts these principles to work. It makes no attempt to replace email or messaging, which between them carry 40+ years of proven utility. Instead it reads your messages, pulls out the tasks, and runs them across the tools knowledge workers already live in.

Beyond GTD: Modern Knowledge Worker Challenge

Getting Things Done, introduced in 2001, still holds up because it speaks to human limits that never really change: we only have so much attention, our memory is leaky, and our priorities compete with each other. Those constraints are about as Lindy-compliant as it gets. They were here before productivity software, and they’ll outlast whatever AI does next.

The trouble is that action items end up scattered across emails, Slack threads, and meeting notes, and that scattering is what this+that calls the “manual tax.” Every minute you spend chasing down a commitment, copying details from one app to another, or hand-updating a task list is friction, and it compounds the more people you have.

Most automation tools won’t do anything for you until you’ve defined the triggers and built the workflows yourself. this+that flips that around: it watches your communication channels and surfaces the tasks on its own, so you’re confirming or tweaking what it found rather than building everything from scratch.

DoBox: AI-Powered Self-Filling Task Manager

DoBox is an AI-fed task manager that fills itself from your inbox. You don’t type tasks in; instead it spots the action items, requests, deadlines, follow-ups, and commitments buried in your incoming messages and adds them for you.

What DoBox automatically captures:

• Action items: Tasks that get assigned outright, or that a message clearly implies
• Deadlines: Time-sensitive commitments that come up in the course of a conversation
• Follow-ups: Things you’ll need to circle back on later, picked up from how the thread is going
• Decisions: Approval requests and open choices that someone needs to resolve
• Commitments: Promises that have been made and now need tracking

Every captured task links straight back to the conversation it came from, so you don’t lose the context the way you do when you retype tasks by hand. A colleague mentions a project you talked through three weeks ago, and the whole thread is right there, no digging through the archive required.

How DoBox Automates Task Capture

What sets DoBox apart from a traditional task manager is Lindy-compliant at heart. Task management has been with us since antiquity, moving from papyrus scrolls to paper lists to digital apps, but the underlying need never goes away. The only thing that really changes is how much friction it takes to capture and organize the work.

Look at software dependency chains and you’ll notice the most durable technologies are the ones acting as infrastructure, not as endpoints. SMTP endures because email apps are built on top of it; SQL endures for the same reason databases lean on it. DoBox treats task capture the same way, as a layer that plugs into your communication tools instead of trying to beat them.

On a team, DoBox makes assignments visible to everyone. A manager can see what’s been delegated, what’s been accepted, and what’s done without chasing anyone for a status update, and each person sees their own commitments without keeping a separate tracker of their own.

Workflows: Streamline with AI Automation

Workflows take task capture a step further, into actually getting things done. You can build automations visually or just describe them in plain language, whichever you prefer.

Core workflow components:

• Triggers: The events that kick a workflow off, whether that’s a message pattern, a scheduled time, or you hitting go yourself
• AI steps: The parts that need actual reasoning, classification, or generated content
• Actions: What runs across your connected tools, like updating a CRM, adding a calendar entry, or firing off a notification

That combination fixes a real gap in older automation platforms. A tool like Zapier is great at deterministic workflows, where the same input always gives you the same output. But the moment a workflow needs judgment, say, deciding whether an email is a sales inquiry or a support request, that kind of automation falls apart.

Craft Workflows with Natural Language

Describing a workflow in plain language lowers the bar to automating anything. Rather than clicking through screens to set up triggers and actions, you just say what you want: “When a customer emails about pricing, create a HubSpot deal and notify the sales team in Slack.”

this+that reads that, drafts a workflow to match, and lets you adjust it before you turn it on. It’s a good fit for anyone who knows exactly what they’re after but doesn’t want to learn the mechanics of automation configuration.

History tends to favor tools that solve a real problem over tools that show off what they can technically do. this+that Workflows is squarely in the first camp: it’s there to turn communication into action, not to flex AI for its own sake.

Unrestricted AI Automation: Model Context Protocol

The Model Context Protocol is what lets this+that’s integrations reach past the pre-built connectors. MCP is a standardized way for AI systems to talk to outside tools and APIs, which means this+that can connect to any service that speaks the protocol.

Currently supported integrations include:

• Communication: Gmail, Outlook, Slack, Microsoft Teams
• Documentation and project management: Notion, Jira
• Development: GitHub
• CRM: HubSpot
• Storage: Dropbox, Google Drive

That’s an architectural bet grounded in Lindy thinking about how long software lasts. If you stack up software dependencies, the apps built on open standards consistently outlive the ones wired to proprietary integrations. Because this+that connects through MCP, when an individual tool changes its API the fix happens at the protocol level instead of forcing a rebuild from scratch.

Connecting Beyond Pre-built Integrations

If your company runs internal tools or specialized software, you can wire them up through MCP servers without waiting for anyone to ship an official integration. That matters in the enterprise, where the standard productivity apps are only a slice of the whole stack.

This is the hazard rate at work in your integration architecture. Tie a platform tightly to a handful of specific tools and it goes down when those tools fade. Build it on open protocols and it can roll with changes in the tool ecosystem without being torn down and rebuilt.

DoBox for Gmail: Chrome Extension

DoBox for Gmail pulls task extraction right into your inbox as a Chrome extension. A sidebar shows the tasks it found in the email you’re reading, and you confirm, edit, or dismiss each one with a single click.

Extension capabilities:

• Real-time task identification: Action items surface while you’re still reading the email
• One-click controls: Confirm, edit, or dismiss a task without ever leaving Gmail
• Workflow triggers: Kick off an automation straight from the message you’re in
• Priority flagging: Flag the items that need attention right now

Going Gmail-first is itself a Lindy-compliant choice. Gmail launched in 2004 and has outlasted 20+ years of competitors, the shift to mobile, and now the move to AI. Meeting people inside an interface they already trust, instead of asking them to adopt a brand-new primary tool, takes a lot of the friction out of getting started.

Real-World Gmail Efficiency Use Cases

Managing client deadlines is a good example of where this pays off. A client emails about a deliverable with a due date, DoBox catches the deadline, and it offers to block your calendar, ping your teammates, or update your project tools, all without you leaving the email.

Approval routing plays out much the same way. An email asks for sign-off on a proposal, a budget, or a decision, and with one click the request goes to the right stakeholders, with a trail back to the original message intact.

Business Process Automation with this+that

The workflow automation examples show how the platform plays out across different business functions. Each one is a common pattern you can drop in as-is or tailor to whatever your situation needs.

Automate Customer Engagement

Customer onboarding automation:

• Trigger: New customer welcome email received or sent
• AI processing: Extract customer details, identify product tier, note special requirements
• Actions: Create CRM record, schedule kickoff call, assign success manager, generate onboarding checklist

Support request routing:

• Trigger: Customer email to support inbox
• AI processing: Classify urgency, identify product area, assess technical complexity
• Actions: Create ticket in support system, route to appropriate team, set SLA timers, notify customer of receipt

Streamline Internal Operations

Meeting follow-ups:

• Trigger: Calendar event ends or meeting notes received
• AI processing: Extract action items, identify owners, determine deadlines
• Actions: Create tasks in project management tool, send summary to attendees, schedule follow-up reminders

Invoice processing:

• Trigger: Invoice attachment detected in email
• AI processing: Extract vendor, amount, due date, line items
• Actions: Create approval request, route based on amount thresholds, update accounting system, schedule payment

The failure statistics make the point that operational efficiency often decides which companies make it through the early years. Automate the repetitive stuff and you free up attention for the work that genuinely needs human judgment, which improves your odds of reaching that Lindy-compliant longevity in the first place.

Who Benefits from this+that?

It earns its keep in roles where the sheer volume of communication turns into coordination overhead.

Engineering leads watch sprint action items scatter across Jira comments, Slack threads, GitHub issues, and email. Once a blocker or a dependency is hiding in a different channel than everyone’s looking at, the timeline slips. Inbox automation gathers the engineering-relevant tasks into one view no matter where they came from.

For sales teams, inbound lead routing is the whole game, and response time moves conversion directly. The instant a prospect emails, the system can spin up a CRM record, alert the account executive, and queue a follow-up sequence, with no manual data entry in between.

Operations heads deal with approval requests landing through every channel at once. Instead of combing email for whatever’s pending, they get the decision points surfaced for them, and completed approvals routed on to the right systems.

Professional services teams have to keep track of client deliverables scattered across project emails, status meetings, and change requests. Pulling out deadlines and setting reminders automatically is what keeps those commitments from slipping through the cracks.

Solving Scattered Work Problem

What ties all these roles together is fragmentation. So many productivity tools have piled up that an important item ends up hiding in whichever one happened to host the conversation. this+that pulls those action items together no matter where they started, giving you a unified inbox for your tasks rather than your messages.

That’s very much in the spirit of the Choose Boring Technology movement in software engineering. Instead of dropping yet another tool on the team to learn and adopt, this+that works with the tools they already have while taking some of the coordination weight off them.

Experience this+that

You can try this+that free and connect your accounts to see how many tasks, and what kinds, are already hiding in the messages you’ve got.

Privacy protections cover the platform. Your email content stays encrypted, and the AI does its processing without anyone at this+that reading your messages.

A Lindy-compliant take on trials matters here too. Platforms that demand commitment up front tend to lose people who bail before they ever see the value. Give someone real time to evaluate the fit and you get a genuine assessment instead of an impulse signup followed by a quick exit.

If Lindy is on your shortlist, it’s worth seeing how the two stack up directly in our this+that versus Lindy comparison, and weighing the wider field in our best Lindy alternatives roundup.

Frequently Asked Questions

How does this+that differ from traditional automation tools?

Traditional automation tools won’t do anything until you’ve spelled out the triggers and actions yourself, which means knowing exactly what you want to automate and building the workflow by hand. this+that turns that around. It watches the communications you already have and picks out the tasks on its own. The AI surfaces what needs doing, you confirm or tweak it, and the system runs it. That works far better for knowledge work, where tasks pop up unpredictably out of conversations instead of following a tidy, repeatable pattern.

What happens to my data if I stop using the platform?

this+that reads your email and message content to find tasks, but it doesn’t keep the message bodies around permanently. Your task data exports in standard formats that other project management tools can read. Data handling follows GDPR and CCPA requirements, and the full policies live in the privacy documentation. You can ask for everything to be deleted whenever you want.

Does this+that work beyond email?

Yes. On top of Gmail and Outlook, it connects to Slack and Microsoft Teams to pull tasks out of chat. The same AI that catches action items in email works across those messaging channels too, so you get one task view no matter where the conversation happened. And as more communication tools adopt MCP, this+that can connect to them as well.

How accurate is AI at identifying tasks?

It leans toward surfacing a possible task rather than letting one slip by, so you can wave off the occasional false positive in a second. Accuracy climbs over time as the AI picks up on what you tend to confirm and what you tend to dismiss in your own style of communicating. Most users tell us the time they spend reviewing flagged tasks is far less than the time they used to lose hunting for action items across channels by hand.

Can teams use this+that without switching tools?

The team features let a manager see delegated tasks and their status without anyone on the team having to give up the tools they already work in. Tasks pulled from email can route into the systems this+that connects to, like Jira and Notion. this+that acts as an extraction and routing layer, so there’s no demand to migrate the whole team onto something new.