Your Criteria Have Changed: How to Update Link Building Evaluation Without Rewriting the Past
A version-control system for evolving link building criteria, preserving the context behind every decision, and reviewing only what genuinely needs attention.
Evaluation Criteria Change Because the Work Learns
Knowing how to update link building evaluation criteria is not about replacing an old list with a new one. It is about managing the evolution of professional judgment without losing the evidence that justified each decision. Criteria are not static. They may change because the brand redefines its priorities, because a campaign pursues a different stage of the funnel, because the team identifies that a signal was being given too much weight, or because editorial, legal, or transparency requirements are added. The availability of opportunities or the way data is verified may also change. Evolving criteria does not mean earlier decisions were wrong. An opportunity approved six months ago may have been reasonable for the objective, budget, available information, and decision framework at the time. Operational quality is not measured by pretending the rules never change, but by recording when they changed and how that change affects outstanding work.
- Treat criteria as a versioned decision framework, not as a sheet that gets overwritten.
- Distinguish between new information, a change in objectives, and a process preference.
- Keep the context of the original decision alongside the criteria version that was applied.
The Problem Is Not Changing the Rules: It Is Comparing Decisions Without Context
When an opportunity researched under an earlier framework is directly compared with another assessed under a recent version, apparent inconsistencies emerge. It may look as though the team has been arbitrary when it has actually applied different rules at different times and for different objectives. The usual reaction is to recalculate the entire history. This is costly and often not very useful. It forces a review of decisions already executed, discards research that remains valid, and turns a process improvement into an endless administrative task. In addition, using the current standard to judge the past can obscure what was actually known at the time. The opposite risk also exists: retaining every previous decision through inertia even when there has been a material change. The answer is selective review. Every change should specify its scope and make it possible to identify which pending, approved, or rejected opportunities may be affected.
- Do not combine comparisons made under different versions without labeling them.
- Do not assume every difference is an error: it may result from a documented framework change.
- Prioritize reviewing open or reversible decisions before actions that have already been completed.
What a Criteria Version Should Include
A useful version should be brief, readable, and actionable. It does not need to become an exhaustive manual, but it must leave enough of a trail to understand what changed and apply the criteria consistently. Assign a simple identifier, such as C-2025-03 or V3. Pair it with an effective date, not only a drafting date. This prevents a rule that did not yet exist from being applied retroactively. Describe the change specifically: it is not enough to say that quality has been refined; explain which requirement, signal, or weighting has changed. Include the reason for the change, the people or roles responsible for approving it, the expected impact on selection, and authorized exceptions. If the change affects paid collaborations, also document the transparency criteria: these collaborations should be identified appropriately, and rel sponsored or nofollow attributes may apply depending on the case and the editorial agreement.
- Identifier and effective date.
- Change introduced and verifiable reason.
- Approval owner and affected teams.
- Expected impact on new, pending, and historical opportunities.
- Exceptions, approval conditions, and planned review date.
Classify the Change Before Deciding Its Scope
Not all changes justify the same response. Classifying them prevents hours being spent reopening records whose meaning has not changed. The first type is documentation adjustments. These clarify a definition, standardize a field name, or add instructions for applying an existing rule more effectively. If the expected evaluation outcome does not change, it is usually enough to use the new wording from its effective date onward. The second type is weighting changes. The opportunity is still assessed using the same variables, but their relative importance changes. For example, a topical relevance objective may be given more weight than a secondary signal. These changes usually require reviewing pending opportunities and those that fell close to a decision threshold. The third type changes vetoes, minimum requirements, or eligibility conditions. This includes a new editorial requirement, a brand restriction, a transparency condition, or a criterion that disqualifies an opportunity that was previously admissible. This is the change most likely to require a priority review.
- Documentation adjustment: changes the explanation, not the expected decision.
- Weighting change: changes the ordering or priority of opportunities.
- Veto or minimum requirement change: can make an eligible opportunity ineligible, or the other way around.
Decide What to Review With an Impact and Status Matrix
Historical review should answer two questions: how much does the change alter the potential decision, and what status is the opportunity in? Together, these variables create a simple, defensible matrix. On the impact axis, classify the change as low, medium, or high. Impact is high when it changes a veto, a minimum requirement, or a decisive factor for the project. It is medium if it can alter priority, indicative budget, or outreach order. It is low if it only improves documentation or affects a marginal signal. On the status axis, separate rejected opportunities, researched opportunities with no decision, shortlisted opportunities, approved opportunities pending execution, and actions already completed. Opportunities without a decision and those approved but still reversible are usually the first priority. Historical rejections are reviewed only when the change could have made them eligible. Completed actions remain with their original context; they can be noted for learning, but there is no point in rewriting their evaluation as though the decision were made today.
- High impact and an open decision: review immediately.
- Medium impact and an active shortlist: review in the next planning cycle.
- Low impact or completed action: retain the record and apply the new version going forward.
- Rejected with a removed veto: review only if it remains relevant to the current project.
The Minimum Record That Provides Traceability Without Bureaucracy
Each opportunity does not need an endless case file. It needs an unambiguous reference to the framework under which it was researched and a note explaining the decision. This record makes it possible to reconstruct the reasoning, compare fairly, and filter which items require review when a new version arrives. At a minimum, link each opportunity to the project, the evaluation date, the version identifier, the decision status, the determining criteria, and the person or role that validated the step. Add a review field: not reviewed, not applicable, pending, reviewed with no change, or decision updated. If the conclusion changes, retain both the previous and new decisions, along with the reason for the change. Project organization tools, metric comparisons, and list analysis can reduce manual work, but they do not replace judgment. SEO metrics are signals that must be interpreted in context; they are not equivalent across sources and do not by themselves prove that an opportunity is suitable.
- Opportunity ID and associated project.
- Research date and criteria version applied.
- Decision status and determining factors.
- Validator, approved exception, and relevant evidence.
- Review status following later changes.
Example: A New Editorial Requirement Without Invalidating Everything Before It
Imagine the team adds an additional editorial requirement for new collaborations: before moving forward, a specific publication and transparency condition agreed by the brand must be verified. This condition is recorded as a minimum requirement from the effective date of version C-2025-03. Opportunities researched before that date do not become poorly evaluated. They retain their original version and evidence. However, those that are shortlisted or approved but not yet executed must be reviewed, because they can still move forward under the new requirement. Opportunities already executed are labeled as evaluated under the previous version; if a useful lesson is identified, it is added to the process record rather than altering the history. It may also be reasonable to review earlier rejections if they were rejected solely because of an aspect that the new requirement makes irrelevant. But this review should be limited to opportunities that are still relevant to current objectives. The purpose is not to recover the entire archive, but to identify open or potentially different decisions.
- Define the date from which the requirement becomes mandatory.
- Review pending or reversible opportunities first.
- Keep completed actions under their original framework.
- Record exceptions explicitly, with the owner and justification.
Frequently asked questions
Do all opportunities need to be reassessed when link building criteria change?+
No. First classify the change and cross its impact with the status of each opportunity. Prioritize open, shortlisted, or approved-but-reversible decisions. Keep completed actions with the context and criteria version applied at the time.
What is the difference between updating a weighting and changing a veto?+
A weighting update changes the relative importance of existing signals and may alter priorities. A veto or minimum-requirement change affects eligibility: an opportunity may no longer be admissible or may require additional validation before moving forward.
How should an exception to the criteria be documented?+
Record the applicable version, the exceptional condition, the reason, the person or role approving it, the scope, and any control measures. An unrecorded exception becomes indistinguishable from inconsistent process application.
Are SEO metrics enough to reassess an opportunity?+
No. Metrics can help guide the analysis, but they must be interpreted alongside relevance, editorial context, project objectives, collaboration transparency, and other documented factors. They are not absolute truths or equivalent across tools.
How should a criteria change be communicated to a client?+
Explain what has changed, when it applies, which pending decisions will be reviewed, and which records will be retained. Present it as an improvement in control and learning, not as a promise of results or an automatic disqualification of earlier decisions.
Sources and references
- Google Search Essentials — Google Search Central
- Spam policies for Google web search — Google Search Central
- Qualify outbound links — Google Search Central