Ahrefs has spent more than a decade selling marketers the data to understand how people find things online. On 12 August it began selling the machine that acts on it. Letaido, the company’s new agent-powered workspace, runs multi-step marketing tasks against Ahrefs’ own web index, and the launch led not with a feature list but with a number: one customer’s account that work which used to fill a 40-hour week now takes an hour.

The repositioning matters because marketing teams are under pressure to hand real work to AI agents at exactly the moment the evidence for doing so has turned cautious. Independent surveys through 2026 have found most teams still trialling agents rather than running them, and a run of research published this month puts hard numbers on disappointing returns. Letaido launches into that gap, and the headline productivity claim it chose to lead with sits on the wrong side of it.

What does Letaido actually do?

Letaido is a shared workspace in which marketing teams hand multi-step jobs to an AI agent that reads Ahrefs data directly, rather than a chat window one person prompts at a time. The agent, which Ahrefs calls Agent A, can plan and run recurring workflows, build dashboards and reports, and monitor a team’s own sites and its competitors, with work continuing on built-in hosting after the conversation that set it up has ended.

It reads Site Explorer, Keywords Explorer, Site Audit and Brand Radar without spending the account’s Ahrefs API quota, drawing on an index Ahrefs says covers more than 170 trillion pages and AI-visibility data from over 439 million monthly prompts. Ahrefs, which describes itself as bootstrapped and profitable on more than $100m in annual recurring revenue, calls Agent A the first of several specialist agents the platform will host.

One worked example the company offers is the board-meeting question of whether a brand appears when a buyer asks ChatGPT about its category. The workspace tracks which brands surface across large language models, flags when a competitor gains ground, and drafts a brief to respond. That capability sits alongside Brand Radar, the AI-visibility product Ahrefs has been expanding through 2026.

Where does the 40-hour-to-60-minute figure come from?

The figure comes from one named customer, not from Ahrefs’ own testing. Ross Simmonds, chief executive of the B2B agency Foundation, is quoted in the announcement saying that his team’s keyword research and bottom-of-funnel content audits, work that once took a full 40-hour week, now take 60 minutes. Ahrefs published no methodology, sample size or customer count behind the claim, which rests entirely on that account and describes a narrow category of task rather than marketing work in general.

A 97.5% cut stands out against comparable claims elsewhere in the sector. Adobe reported marketers saving 114 minutes a week on content creation in 2025, and Smartly’s PwC-validated figure put savings at 42 minutes recovered per hour of work. Set against those, a single customer’s account of compressing a week into an hour is the outlier, not the benchmark.

The distinction worth holding is the ordinary one between what a vendor claims and what it shows. Simmonds’ account may prove out under wider use; it has not yet been demonstrated outside the company that stands to gain from it, and marketing teams weighing Letaido have only the one quote to go on.

What does Letaido cost, and what is it launching into?

Letaido costs £73 ($99) a month, a flat subscription that includes one workspace, £37 ($50) of AI credits and unlimited team members. The credits are charged as real dollars against model and connector calls, with roughly a quarter of each one kept by Letaido as a service fee and the rest passed to the model provider. Larger workspaces run up to £124 ($169) a month.

One condition matters for the productivity pitch. The agent’s access to Ahrefs data mirrors the account’s existing Ahrefs plan, so a team that wants the full index behind those workflows still needs a paid Ahrefs subscription underneath the £73 fee. Connect a free Ahrefs account and the data the agent can pull is limited to what a free plan allows.

The wider market gives marketers reason to test the promise before banking on it. KPMG’s second-quarter research found 49% of organisations had scaled back agentic AI rollouts because costs outran the value returned. TransUnion reported in August that only 53% of marketing leaders saw meaningful returns on their AI spending, while 65% still measured its impact through estimated time saved rather than controlled tests. A Typeface survey the same summer found campaign timelines had lengthened, not shortened, under AI, with approval bottlenecks the main brake. Ahrefs is not alone in the move: Salesforce made autonomous agents standard across its marketing platform and priced them by the action earlier this year.

Ahrefs published pricing and a customer quote, but no independent data behind the figure that led its announcement. The research it will be measured against is already public.