The SEO Scorecard
The SEO scorecard tracks whether organic search is sending more of the right people to your site and whether they do something when they land. It pulls from Search Console, your analytics tool, and your publishing calendar, and the marketing lead runs it with the web developer and content lead each owning rows. Twenty minutes a week, every number owned.
Why this scorecard exists
Why most SEO reports never change a decision
A generic marketing scorecard puts one row on SEO, usually organic sessions, and calls it done. That row moves for reasons nobody in the room controls: a Google update, a seasonal swing, a brand campaign on another channel. When it dips, the answer is a shrug. When it rises, everyone takes credit. A number nobody can act on is decoration.
Organic search moves slowly and the causes are spread across three people. Content ships a page, the developer controls whether it gets indexed and loads fast, and the marketing lead decides which pages deserve the links. The scorecard has to separate inputs you control this week (pages shipped, pages indexed, links added) from outcomes that lag by months (position, clicks, conversions). It also has to split brand from non-brand, because brand clicks tell you about your reputation and non-brand clicks tell you about your SEO.
Once the scorecard is live, the weekly conversation stops being about traffic and starts being about the specific thing that broke. Indexing dropped, so the developer checks the sitemap. A money page slid, so the content lead adds two internal links. AI-assistant referrals showed up for the first time, so someone checks which page they landed on and why. The red-flag rules mean nobody waits for a quarterly report to notice.
The scorecard
10 numbers. Every one has an owner.
| Metric | What it means | How to set the target | Owner | Cadence | Red flag |
|---|---|---|---|---|---|
| Non-brand impressions Impressions rise weeks before clicks do, so this is your earliest signal that new content is being seen. | How many times your pages appeared in Google results for searches that did not include your brand name, from Search Console with brand terms filtered out. | Use your trailing 12-week median as the floor and set the target as that median plus the impressions you expect from the pages shipping this quarter, then revisit once those pages are indexed. | Marketing lead | Weekly | A week-over-week drop larger than your worst normal weekly swing in the last 12 weeks triggers a manual-actions and indexing check the same day. |
| Non-brand clicks This is the number SEO work is supposed to move, and separating it from brand keeps a good PR week from hiding a bad SEO month. | Visits from Google search where the query did not contain your brand name, from Search Console. | Start with your trailing 8-week median, then set each quarter's target as last quarter's actual plus a percentage you choose based on how many money pages are still off page one. | Marketing lead | Weekly | Two consecutive weeks below target starts a query-level review of the ten pages that lost the most clicks. |
| Non-brand share of clicks If share is falling while total traffic holds, brand demand is masking an SEO decline and nobody will see it in the top-line number. | Non-brand clicks divided by total organic clicks, expressed as a percentage. | Record the share from your last full quarter and set the target to hold or grow it, and if you run brand campaigns on other channels, expect share to dip during those weeks and note it on the board. | SEO lead | Monthly | Share falling two months in a row while brand clicks stay flat means the decline is real and moves to the top of the next weekly review. |
| Pages indexed vs published A page that is not indexed cannot rank, and indexing problems are the most common reason new content produces nothing. | The count of pages Google reports as indexed in Search Console compared to the count of live pages you intend to rank. | The target is every intended page indexed, so set it equal to the published count and treat the gap as the number to watch, not the ratio. | Web developer | Weekly | Any page still unindexed 14 days after publish, or any drop in the indexed count with no matching removals, goes to the developer that day. |
| Average position, money pages Site-wide average position is noise, while the position of the twenty pages that make you money is the whole game. | The average Google ranking position for the pages that drive leads or sales, from Search Console, tracked for each page's main query. | List the money pages, record each one's current position for its main query, and set the target as one full position better per quarter for anything ranked below the top three. | SEO lead | Weekly | Any money page slipping more than three positions in a week gets a same-week check for a competitor change, a lost link, or a technical error on the page. |
| Core Web Vitals pass rate Failing pages carry a ranking handicap and convert worse, and this is the one row the developer fully controls. | The share of your URLs that Search Console marks as Good on all three Core Web Vitals, the page-speed and visual-stability measures Google uses. | Pull the current pass rate from the Core Web Vitals report, set the target to get every money page to Good first, then the rest of the site template by template. | Web developer | Monthly | Any money page moving from Good to Needs Improvement or Poor triggers a template review before the next release ships. |
| New content shipped It is the input you control most directly, and a scorecard with only lagging outputs gives the team nothing to do on Monday. | Count of new or substantially rewritten pages published this week against the content plan. | Set it from your actual publishing capacity over the last eight weeks, not from what you wish it were, and raise it only after the team has hit it four weeks running. | Content lead | Weekly | Two weeks of zero against a plan that called for pages means the plan or the capacity is wrong, and the review decides which. |
| Internal link coverage Internal links are the cheapest ranking lever you own and the one most often forgotten after a page ships. | The share of money pages that have at least the minimum number of internal links from relevant pages that you set for your site. | Set a minimum link count per money page that fits your site size, audit current coverage once, and set the target as full coverage by the end of the quarter. | Content lead | Monthly | Any new money page still below the minimum link count 30 days after publish gets its links added before any new content ships. |
| AI-assistant referral sessions More research now happens inside AI answers, and this row tells you whether your pages are being cited and clicked, which is what GEO, generative engine optimization, is trying to change. | Sessions that arrive from AI assistants such as ChatGPT, Perplexity, Gemini, and Copilot, identified by referrer in your analytics tool. | Start with a raw count and a landing-page list, since most sites begin near zero, and set the target as month-over-month growth once you have three months of data to compare. | Marketing lead | Weekly | A month with zero AI referrals after a month with some prompts a check that the referrer sources are still being captured, then a review of which pages lost their citations. |
| Organic conversions Traffic that does not turn into business is a cost, and this row is the one the owner reads first. | Leads, calls, form fills, or purchases that started from an organic search visit, from your analytics tool or CRM. | Take your trailing 8-week median of organic conversions and your current organic conversion rate, then set the target from the non-brand clicks target above times that rate so the two rows agree. | Owner | Weekly | Conversions falling while clicks hold means the landing pages or the forms broke, and the developer checks tracking and forms before anyone touches content. |
Targets are set from your own history, never from someone else's benchmarks. The builder does the arithmetic once you plug in your numbers.
The weekly review
Twenty minutes, same order, every week.
// 01
Organic conversions first
Open with the number the business feels. Did organic produce the leads or sales it was supposed to this week? If yes, the rest of the review is maintenance. If no, the next three steps find out why.
// 02
Split traffic from tracking
Put non-brand clicks and impressions next to conversions. Clicks holding while conversions drop points to a tracking or form problem, not an SEO problem. Clicks dropping points to indexing, position, or a Google change, and the developer and SEO lead each speak to their rows.
// 03
Inputs: what shipped and what got indexed
The content lead reports pages shipped against plan and link coverage on anything new. The developer reports the indexed count and any Core Web Vitals movement. This is the part of the review that is about next week, not last week.
// 04
Decide one thing
Every red flag on the board gets an owner and a date before the meeting ends. If nothing is red, the decision is which money page gets the next content or link push. The marketing lead writes it down and it opens next week's review.
Where scorecards die
Mistakes we see constantly.
Reporting total organic sessions
Total sessions mix brand searches, non-brand searches, and returning customers into one line that moves for reasons SEO did not cause. Split brand out or the scorecard will tell you the wrong story.
Tracking site-wide average position
One number across every query on the site drowns the pages that matter in long-tail noise. Track the money pages and their main queries instead.
Skipping the indexing row
Teams publish for months before noticing half the pages never made it into Google. Indexed versus published is the row that catches it in the first week.
Only lagging metrics
A scorecard with nothing but clicks and conversions gives the team no lever to pull on Monday. Content shipped and links added are the rows that turn it into a plan.
Reacting to one bad week
Organic search is noisy week to week and slow to respond to changes. Set red-flag rules on consecutive weeks or on drops larger than your normal swing, and let a single bad week pass.
Other scorecards
Same discipline, different numbers.
By platform
By business type
Want it set up live in your dashboards?
We install scorecards on real data, wire the owners and alerts, and coach the weekly review. Scorecards and accountability is the service.
Questions we get
Straight answers.
How do I separate brand from non-brand in Search Console?
Use the query filter to exclude your brand name and its common misspellings, and save the filtered view so the numbers are consistent week to week. Check the filter twice a year, because brand terms drift as people learn new product names.
How often should we look at SEO numbers?
Weekly for the scorecard, with the understanding that most rows move slowly. The weekly rhythm exists to catch breakage fast, indexing drops, tracking failures, and position slides, not to judge the strategy every seven days.
What counts as a money page?
Any page whose job is to produce a lead or a sale, usually service, product, and location pages, plus the two or three articles that reliably feed them. Keep the list under twenty so the row stays readable.
How do we track AI-assistant referrals?
Most analytics tools show them under referral traffic with sources like chatgpt.com or perplexity.ai. Build a channel group or segment for those sources so the row updates on its own, and review the source list quarterly as new assistants appear.
Who owns the SEO scorecard if we have no SEO person?
The marketing lead owns the board, and the developer and content lead own their rows. The SEO lead rows go to the marketing lead until someone fills that seat, and the scorecard makes the gap visible instead of hiding it.
A scorecard is only as good as the system feeding it.
Start with a diagnostic call. We will tell you which of these numbers you can trust today and what it takes to trust the rest.