Your association's website went live in 2018 and has been added to ever since. Nobody has removed anything. Search traffic has been sliding for about three years, slowly enough that no single month ever triggered a conversation, and last week someone on the board finally asked why the organization does not come up on Google anymore. The proposals you are about to receive will all recommend a redesign, and most of them will include a line item called SEO refresh.
Some of that work is genuinely worth doing. A meaningful share of what will be recommended to you is not documented by Google at all, and one or two items are things Google explicitly identifies as the wrong instinct. The difference is worth about the cost of the project.
How we checked this. Everything attributed to Google below is from Google Search Central's own documentation, checked this month, and quoted rather than paraphrased. Where something is common practice that Google has never published, we say so. That distinction is the whole point of this post.
Three Things You Will Be Told That Google Does Not Say
These come up in almost every proposal we see reviewed. All three are widely believed. None is Google's stated position.
- "Delete your thin old pages and the rest of the site will rank better." Google's guidance on core updates says deleting content is a last resort, and only to be considered if you think the content cannot be salvaged. Its self-assessment for helpful content asks, as an example of search-engine-first behavior, whether you are removing a lot of older content primarily because you believe it will help your rankings overall. Consolidating and improving weak pages is documented. Deleting them to lift everything else is not.
- "Update the dates on your old posts to keep them fresh." Google's freshness systems are described as query deserves freshness. They apply to queries where fresher content is expected, not as a site-wide recency bonus. Your membership benefits page is not such a query. And the same self-assessment asks directly whether you are changing the date of pages to make them seem fresh when the content has not substantially changed.
- "Add schema markup so you show up in AI Overviews." Google's guidance on AI features is unusually blunt: there is no special schema.org structured data that you need to add, and there are no additional requirements to appear in AI Overviews. Structured data is worth adding for rich results and for helping Google understand your pages. It is not the entry ticket to AI answers that it is being sold as.
What Google Does Document
The list of things Google actually confirms is shorter and less exciting than the pitch, which is exactly why it is useful for judging a proposal.
Core Web Vitals are a real ranking input. Three metrics, with published thresholds: Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint at 200 milliseconds or less, and Cumulative Layout Shift of 0.1 or less. Note that Interaction to Next Paint replaced First Input Delay in March 2024, so a report built on the old metric is out of date.
But relevance beats speed, in Google's own words. Google Search always seeks to show the most relevant content, even if the page experience is sub-par, and a great page experience contributes in cases where lots of helpful content is already available. A fast page about nothing does not outrank a slow page that answers the question.
HTTPS and mobile-friendliness are no longer independent ranking boosts. This retires a decade of proposal boilerplate. Google states that beyond Core Web Vitals, other page experience aspects do not directly help your website rank higher in search results. Mobile still matters, but through mobile-first indexing, Google indexes the mobile version of your content, rather than through a mobile-friendly bonus. Google retired the Mobile Usability report and the Mobile-Friendly Test in December 2023.
Near-duplicate pages are a canonicalization problem, not a penalty. Keyword cannibalization is a search-industry term, not a Google one, and there is no documented cannibalization penalty. What Google documents is how it picks one canonical URL among similar pages, and the relative strength of your signals: a redirect is strong, a rel=canonical tag is strong, inclusion in a sitemap is weak, and the methods stack.
Why an Association Site Ages Differently
Associations do not have an old website problem. They have an accumulation problem. A business rebuilds and abandons the previous site. An association keeps the 2019 conference program because a member might need it, keeps the old chapter pages because the chapters asked, keeps the retired program page because it might come back, and keeps eleven years of board minutes as PDFs. Nothing is removed, because removing anything requires someone to have authority to say a thing is finished.
That produces a specific shape of problem. Six near-identical membership pages written by six different staff members over eight years, all competing to be the canonical answer for the same query. A conference microsite outranking the current conference. A 2019 fee schedule that a member finds before the current one.
And this is a real user-facing failure, not just a ranking one. When a member calls to complain that the fee on your website is wrong, the fee on your website is wrong. It is on a page you forgot existed, and Google is quite reasonably showing it to people because you never told it otherwise.
One Thing Worth Checking Before You Buy Anything
Crawl budget almost certainly is not your problem. This gets raised in proposals for sites of every size, and Google scopes it tightly: the guidance is aimed at sites with over a million pages that change weekly, or over ten thousand pages that change daily. Google says plainly that if your pages seem to be crawled the same day they are published, you do not need to read that guide at all.
There is one exception that does apply to smaller sites, and it is worth ten minutes in Search Console. If a large share of your URLs sit under Discovered, currently not indexed, that is a genuine signal worth investigating. It is also free to check, and it tells you something a vendor cannot tell you from the outside.
While you are in there, confirm your pages are actually eligible. For AI features specifically, Google documents two requirements and they are both binary: the page has to be indexed, and it has to be eligible to appear with a snippet. A nosnippet directive left over from an old build will quietly exclude you, and that is a five-minute fix rather than a project.
The Work That Actually Helps
On an accumulated association site, in the order we would do it.
- Find the duplicates and pick a winner. Not deletion, consolidation. Choose the page that should rank, and point the others at it with redirects and canonical tags. This is the documented mechanism, and on most association sites it is the single highest-value piece of work.
- Fix the pages that are wrong rather than removing them. Old fee schedules, retired staff, superseded guidance. Google's position is that improvement comes before removal, and it happens to be what your members need too.
- Measure Core Web Vitals on the pages that matter. Not the home page. The membership page, the events page, the donation or dues page. Those are the pages where a ranking input and a conversion problem are the same problem.
- Retire what is genuinely finished, without expecting a ranking lift. A 2019 conference page with no traffic can be redirected somewhere useful because it is confusing to members, not because subtraction improves the rest of the site. Do it for the right reason and you will not be disappointed by the result.
- Check what happens if URLs change. If any of this involves moving pages, Google says to keep redirects for at least a year so ranking signals transfer, and warns to expect fluctuation while it recrawls. It gives no promised recovery timeline, and neither should anyone quoting you.
How Long Any of This Takes
Worth setting expectations with a board before the work starts rather than after. Google folded its helpful content system into core ranking in March 2024, and its guidance on recovery is candid. Some changes can take effect in a few days, but it could take several months for Google's systems to confirm a site is now producing helpful content. If nothing has happened after a few months, that could mean waiting until the next core update.
Google runs roughly two or three core updates a year. So the honest version of the timeline is months, not weeks, and anyone promising a rankings recovery by a specific date is describing something Google does not offer.
When a Redesign Is the Right Answer Anyway
None of this is an argument against rebuilding. If your site cannot be edited without a developer, if it is on an unsupported platform, if it fails accessibility badly, or if the content model cannot express what your association now does, rebuild it. Those are all good reasons and they stand on their own.
The argument is narrower: a redesign does not fix SEO by itself, and a redesign that migrates the same eight hundred pages into a new template carries every duplicate and every wrong fee schedule across with it. The content work is the SEO work. The visual refresh is a separate thing you may also want.
And if your traffic is stable, your pages are indexed, and members find what they need, an old website that works is not a problem requiring a solution. Plenty of association sites are older than they look and performing fine.
Which of Your Pages Are Competing
Point us at your website. We will map every page Google has indexed, group the ones that are competing for the same queries, flag the pages carrying information that is no longer true, measure Core Web Vitals on the pages that actually convert, and check whether anything on your site is excluded from search or from AI answers by a leftover directive. You will get back a written list with a specific recommendation next to every line, consolidate, improve, redirect or leave alone, and an honest note about which items Google documents and which are the industry guessing.