If 40 hours of research can be finished in an hour, can our team do it too?
For teams spending days on keyword research and content audits, this is a hard number to ignore. Ross Simmonds of content agency Foundation said that after using Ahrefs’ AI marketing agent workspace Letaido, keyword research and bottom-of-funnel content audits that had taken 40 hours a week were completed in 60 minutes.
But when you consider a purchase, you also encounter research that seems to point the other way. In a Gartner survey, 45% of martech leaders piloting or implementing AI agents said agent capabilities from existing vendors had failed to meet their promised business-outcome expectations.
These two numbers are not a scorecard separating winners from losers. One is a specific customer’s work case; the other is a survey spanning multiple vendors. The decision to adopt Letaido should be based less on “which number is right?” and more on whether time savings can be reproduced for the same scope of work while accounting for quality and cost.
A 97.5% reduction is calculable, but the average effect is unknown
Based only on the stated work time, the reduction from 40 hours to 1 hour is 39 hours, or 97.5%. But the Foundation case does not disclose the site and keyword count, source inputs, quality acceptance criteria, human review and editing time, or AI credits consumed. That means this figure cannot be generalized as Letaido customers’ average savings rate or an entire marketing team’s productivity gain.
The product’s work scope itself is documented in official materials. Letaido handles multi-step tasks and creates reports and dashboards, while retaining files, databases, scheduled jobs, and persistent memory in the workspace. It can also connect Ahrefs data and external services. However, the scope of capabilities described by a provider does not guarantee output accuracy or the reliability of unsupervised operation.
Read 40 hours→60 minutes as a hypothesis to test, not a purchasing target. Even if agent execution time shrinks, total time may not fall as expected if people spend a long time checking evidence and fixing errors. Subscription and request costs must also be included to see the actual unit cost of the work.
45% is neither Letaido’s failure rate nor a work-time increase rate
Gartner surveyed 413 marketing technology leaders from June through August 2025. Of all respondents, 81% were piloting or implementing AI agents, and 45% of that group said agent capabilities from existing vendors fell short of promised business-outcome expectations.
| Figure | Verified subject | What it cannot be generalized to say |
|---|---|---|
| 40 hours→60 minutes | One Foundation team’s keyword research and bottom-of-funnel content audit | Average savings rate across all Letaido customers |
| 45% | Share of respondents piloting or implementing agents from multiple vendors whose expectations were unmet | Letaido failure rate or share of users whose time increased |
So saying, “one team saved 39 hours, while 45% experienced the opposite,” mixes denominators and performance criteria. Gartner’s result can serve as a warning that a large time-saving case does not automatically guarantee business outcomes, but it is not evidence for judging Letaido’s own success or failure.
Beyond $99 per month, per-run costs are visible separately
According to current official documentation, Letaido’s base subscription is $99 per month. It includes one Medium workspace, $50 in AI credits each month, and unlimited team members. Additional credits, larger workspaces, or additional workspaces may incur separate charges.
Connecting Ahrefs endpoints is included with a Letaido subscription and does not consume Ahrefs API units, but the SEO data the agent can read follows the scope of the connected Ahrefs plan. You need a plan that supports the necessary projects, reports, and SEO data, as well as read permission.
You can review run costs in the workspace’s Controls, broken down into LLM Requests, HTTP Requests, and API Calls. Since these records do not include human review time, the cost screen alone cannot calculate ROI.
Start with one audit you completed in the past
Foundation’s exact conditions are not public, so they cannot be reproduced as-is. Instead, choose one site audit with a recorded input scope and manual-work history to create a comparable record from the start of adoption.
- Check the sign-up terms first. At letaido.com, select the current Get Letaido button. Depending on the screen, it may appear as Subscribe, while the official quick-start guide refers to this entry button as Get started. Sign in with a Letaido account separate from your Ahrefs login, then complete payment for the $99 monthly subscription. If you are an existing Ahrefs user, check before payment whether your Ahrefs dashboard actually shows a dedicated one-month trial invitation worth $99; apply it only if it is shown. A trial is not guaranteed for general new sign-ups.
- Create a workspace. Choose an organization name and a subdomain to use as the workspace address, then confirm the included Medium workspace’s name and region before creating it. You can defer inviting the team until after the first run.
- Connect Ahrefs only when needed. For an audit that needs SEO data, select Connect next to Ahrefs and authorize Letaido to read projects, reports, and SEO data. Also confirm that the necessary data is included in your current Ahrefs plan.
- Confirm the screen is ready. Open the workspace tile. If you see the chat panel on the right and Apps, Downloads, Connectors, Controls, and Settings on the left, you can begin the first prompt.
- Request the same scope as the prior audit. For example, you could enter: “[Pilot example] Audit the SEO health of example.com and summarize the 10 highest-impact issues in a table. For each item, include supporting data, affected URLs, priority, and recommended action. Before running, confirm the required site scope and any external actions that will incur costs.” The 10 issues and output fields are examples for comparison, not results guaranteed by Letaido.
- Record results and costs together. After the agent asks clarifying questions or begins work, have a person review the supporting data and affected URLs. Record the time spent fixing omissions and errors, along with the Controls costs for LLM Requests, HTTP Requests, and API Calls, next to the existing manual-work record.
Work and input scope: ______
Previous total manual time: ______
Agent execution time: ______
Human review and editing time: ______
Internal quality pass status and evidence: ______
Request costs: ______
Next decision: continue / revise scope / stop
The first success criterion is not filling all 10 example issues or reproducing 97.5%. Once you have reviewed outputs for the same scope and fully recorded agent costs, human editing time, and internal quality-pass status, you have evidence for deciding on the next run.
Research and execution scope
This article was written after reviewing Letaido’s official launch materials, product overview, quick-start and pricing documentation, the current official landing page, and Gartner’s official announcement. Account creation, payment, an actual site audit, and credit consumption were not performed directly.
The Foundation case’s input scope, review time, and costs were not disclosed. Subscription pricing, taxes, additional-credit terms, and trial availability should be checked again on the official screen at the time of sign-up.
.png)
