Discussion starter: an agency question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- Agency
Is there any SEO guarantee you would put in a contract that is both honest and genuinely useful to the client?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
Yes, but only guarantees about things the agency controls: how it works, what it delivers and when, what the client owns, and what happens if the agency breaks the rules. Those can be honest and genuinely useful, because they protect the client against the ways SEO engagements actually fail. Guarantees on rankings, traffic or revenue are a different category. Google’s own hiring guidance states plainly that “No one can guarantee a #1 ranking on Google”, and outcome guarantees tend to push agencies towards whatever easy wins satisfy the clause.
The test I would apply to any clause has three parts. The agency controls it. The client can verify it without relying on the agency’s own report. A breach has a defined remedy. A clause that fails any of the three is marketing copy, not a guarantee.
Why outcome guarantees fail that test
Rankings and traffic depend on factors the agency does not control. Google shipped announced core updates in March and May 2026 alone. AI Overviews change click-through rates on queries where positions have not moved. Demand shifts with the economy and the season. And on most engagements, the client’s developers decide when recommended changes go live, so the agency cannot even guarantee its own work reaches production.
The bigger problem is incentives. A clause is a target, and targets get met in the cheapest available way:
- “Page one for 10 keywords” gets met by choosing low-competition, low-value keywords, sometimes branded ones the client already ranked for.
- “Traffic up 30%” gets met with informational content that brings visits and no enquiries.
- “20 links a month” gets met with cheap links, which is how clients end up with a spam problem.
There is also legal exposure. In many markets, promising an outcome you know you may not deliver raises misleading conduct questions under consumer law (the Australian Consumer Law is one example). That is a reason to have a lawyer review the wording, not something to settle in a proposal template.
Important distinction:
A forecast and a guarantee are different documents. A forecast with stated assumptions and a range (“if these 12 templates ship by March, non-brand clicks to category pages could plausibly rise between X and Y”) is honest and useful. Google’s hiring guide even suggests asking an SEO what results to expect and in what timeframe. Put forecasts in the proposal and review them quarterly. Keep them out of the guarantee clause.
The guarantees I would put in a contract
Guarantee What it commits to Remedy if breached Failure it protects against Search Essentials compliance No tactics listed in Google’s spam policies (bought links, link schemes, cloaking, scaled low-value content). Every link placement and outreach target disclosed to the client. If a manual action results from the agency’s work: remediation at the agency’s cost and a refund of fees for the affected period Hidden tactics that leave the client with a penalty after the agency has gone Ownership and access Client owns every account (Search Console, analytics, tag manager), all content, briefs, data and tool configurations. The agency works through user access, never as sole owner. Immediate handover on request; termination right Accounts held hostage; lock-in Named deliverables by date Specific outputs per period (audit by week four, a set number of briefs per month, technical tickets written to an agreed standard with acceptance criteria) Pro rata fee credit when the agency misses a deliverable it controls A retainer that buys activity nobody can point to Incident response times Critical issues (site deindexed, robots.txt blocking, a release breaking templates) acknowledged within an agreed number of business hours and diagnosed within an agreed window Fee credit Slow reaction during the incidents that cost the most Reporting integrity Metrics and baselines defined at kickoff, the same definitions every month, raw data access for the client, and losses reported as prominently as gains Audit right; termination for breach Vanity metrics and moving goalposts Clean exit Notice period no longer than 30 days after the initial term; handover of documentation, change log, briefs and link records within a set number of days Withheld final invoice until handover is complete Knowledge walking out the door Two lighter commitments are worth adding. A change log: every change the agency makes to the site is recorded with date and URL, which is what lets anyone diagnose a drop later. And disclosure of who does the work: subcontractors, offshore teams, and which tasks use AI tools with what human review.
Watch for this:
A link “replacement guarantee” (if a link disappears within 12 months, it gets replaced) usually signals that the links were bought. An agency can only promise that a placement stays live if it has a commercial arrangement with the publisher. On a contract, that clause is a red flag, not a benefit.
If the client wants the agency to carry some risk
Some clients reasonably want the agency to share the outcome. There are honest ways to do that, and they rarely take the form of a guarantee.
Structure Honest? Main trade-off Fee at risk on agency-controlled deliverables Yes Protects against non-delivery, but does not tie the agency to business outcomes Upside-only performance bonus on a defined outcome against an agreed baseline Yes, if the definitions are tight Aligns incentives; needs agreed attribution rules and tolerance for time lags Outcome guarantee with a refund Rarely Invites gaming and disputes, and the agency prices the risk into the fee anyway Conditional outcome commitment (“if you ship X by Y, and barring algorithm changes…”) Technically Needs so many exclusions that clients read it as weasel words If I were structuring a bonus, I would define the metric by rule, not by a label a vendor can change. Use non-brand organic conversions from the CRM, with brand defined by an agreed pattern (Search Console’s branded queries filter can inform it), compared against the same months of the baseline year. Channel labels move: GA4 added an “AI Assistant” default channel group in May 2026, so sessions from recognised assistants now land in a channel that did not exist before. A bonus written against a channel label can change value without anyone doing anything differently. The contract should also say what happens when tracking breaks or the client changes the conversion definition.
Hypothetical example:
A mid-sized retailer is migrating ecommerce platforms and asks for a “no traffic loss” guarantee. An honest version commits to what the agency controls: a redirect map covering every URL that earned organic clicks or external links in the past 12 months, a pre-launch crawl of staging against an agreed checklist, and 90 days of post-launch monitoring with defined response times. It adds a forecast range for the dip and recovery, with the assumptions written down, and a named rollback trigger agreed with the client’s developers. It does not promise traffic retention, because the agency does not control launch timing, template changes made after sign-off, or what Google does in the weeks after launch.
How the guarantee set changes by engagement type
- Audit only: scope, delivery date, prioritisation method and a walkthrough session. Compliance guarantees barely apply.
- Technical retainer: ticket quality standards, response times and change logging matter most.
- Content: brief and publishing cadence, subject-matter review, originality, and disclosure of AI use.
- Digital PR and links: full disclosure of every placement, no paid links passing ranking credit, and the client’s right to veto targets.
- Migration: redirect coverage, pre-launch testing, monitoring window and rollback criteria.
The decision rule I would use when reviewing any SEO contract: if a clause names a Google outcome, cut it or move it into the forecast. If it names something the agency controls, the client can check independently, and a breach costs the agency something, keep it. A client who compares two agencies on those terms learns far more about how each one works than a ranking promise could ever tell them.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.