Skip to content

Data for SEO: The Numbers That Should Drive Your Strategy and Where to Get Them Free

Most SEO arguments end the moment someone opens the right dataset. The right datasets are mostly free, and most sites have never joined them.

John Cravey with AIFounder5 min readUpdated Aug 6, 2026

Data for SEO is not a tool subscription. It is four questions answered by four sources, three of them free: what do you rank for today, what does that traffic do for the business, where is demand going, and how big is the market underneath it. Sites that can answer all four make different, better decisions than sites running on a tool's opinion. Here is each layer, what it answers, and how to join them.

Free estimate · 2 minutes

Your Search Console data already shows the gap.

The estimate doubles as intake for an audit: we read your data, show where the clicks are being lost, and price the fix.

The plain-English version

SEO produces numbers constantly, and almost all of them are noise. Rankings wobble, traffic fluctuates, tools disagree. The signal lives in a small set of joins: queries against positions over time, landing pages against conversions, demand direction against your content calendar, and market size against your growth ceiling. Build those four reads once and every SEO argument in your business gets shorter, because someone can look.

Layer 1: query data, the observed truth

Google Search Console is the one dataset in SEO that is measurement rather than estimate: real queries, real impressions, your real average position. It answers: what do we rank for, where are we close, what is slipping. The two highest-value reads are near-miss queries (positions 8 to 20 with impressions, your fastest wins) and the decay watch (pages whose clicks erode quarter over quarter). Setup to first insight is an afternoon: GSC from zero to actionable, and the discipline of reading it without fooling yourself is its own skill.

GSC now also carries the AI-era read: how often you appear in AI Overviews and AI Mode, which behaves differently from classic results. We covered that workflow in measuring AI search visibility.

Layer 2: outcome data, or whether any of it paid

Traffic is not the product. GA4 joins organic sessions to what visitors actually did: calls, forms, purchases, by landing page. The configuration that matters is a short list of real conversion events, covered in the GA4 conversion events piece, and the highest-value view is the GSC-to-GA4 join: query family to landing page to conversion. That join is what tells you a page ranking third for a converting term deserves the next unit of work more than a page ranking first for a term that pays nothing. The full reporting view is in joining GA4 with Search Console.

Layer 3: demand data, direction and season

Keyword Planner supplies volume ranges; Google Trends supplies the two things volume hides: trajectory and seasonality. A term at index 40 and climbing beats one at 70 and sinking, and every seasonal business needs content indexed a quarter before its season, because rankings need lead time. Trends is also the cheapest reality check on a tool's volume claim: a term a tool calls big that flatlines in Trends deserves suspicion. Treat all demand numbers as ranges with a direction, never as facts.

Layer 4: market data, the ceiling underneath the searches

Search demand floats on real-world demand, and the real world is measured, free, by the public statistical system. County Business Patterns counts business establishments by industry and county: how many HVAC firms, law offices, or dental practices exist in a market, which is both your competition count and, for B2B, your prospect count. The American Community Survey describes the households: incomes, housing stock, growth. This is the layer that answers strategy questions keyword tools cannot touch: which metro to expand into, whether a segment is big enough to justify a content line, what share of a market you already hold. It is the Census layer of how we scope every engagement, and it is why our market reads start there rather than in a keyword tool.

Joining the layers: the monthly read

  1. Pull GSC: near-miss queries, decaying pages, AI-surface impressions. Pick the three moves they imply.
  2. Pull the GA4 join: conversions by organic landing page. Confirm the pages you are investing in are the ones that pay.
  3. Check Trends for your top ten terms quarterly: direction, season, anything breaking out that deserves a page.
  4. Revisit the market layer twice a year: has the census picture (new households, new businesses) changed what is worth targeting?

That is the whole system: an hour a month, four sources, decisions grounded in the strongest data available at each altitude. The keyword-selection process that consumes these inputs is in how to find keywords for SEO, and the execution playbook is in SEO for small business.

Where the paid suites fit

Rank trackers and all-in-one SEO suites are aggregation conveniences, not additional truth: everything they show is modeled from the same public surfaces, plus daily rank checks the free stack does not give you. They earn their fee when you manage many sites, need competitor rank history, or want alerting without building it. They do not earn it as a substitute for the four layers above, and their volume numbers deserve the same distrust as any estimate. Buy tooling when the free stack's manual hour becomes the bottleneck, not before, and keep decisions anchored to the observed layers (Search Console and GA4) no matter whose dashboard displays them.

Keep a change log, or the data lies to you

One discipline turns all four layers from trivia into evidence: write down what you changed and when. A dated one-line log (rewrote the water heater page, launched the review push, migrated hosting) is what lets you read a traffic move as cause and effect instead of weather. Without it, every ranking gain gets credited to whatever someone remembers doing, and every drop starts a hunt with no map. The log plus the monthly read is the entire operating system; it fits in a spreadsheet and it is the difference between running SEO on data and narrating it after the fact.

What this looks like at the next level

  • One site, one owner: the monthly read above, in an hour, from the four free sources. No dashboard needed; the discipline is the dashboard.
  • Multi-location or multi-service: the same reads split by location and service line, because an average across locations hides the one that is slipping.
  • Agencies and in-house teams: the joins automated (GSC and GA4 both export to spreadsheets and warehouses), thresholds alerting on decay, and the change log kept per property. Same four layers, standing infrastructure.

Want the four-layer read done for your business? Run the estimator and we will pull your market's census read and search demand before any sales conversation. It is the first phase of how every engagement works.

Answers

Frequently asked questions

What data do I need for SEO?

Four layers: query data (Google Search Console: what you rank for, impressions, clicks, position), outcome data (GA4: what organic visitors do and convert to), demand data (Keyword Planner and Google Trends: volume and direction), and market data (public census sources: how many businesses or households in your area could need you at all). Each answers a question the others cannot.

Is Google Search Console data enough for SEO decisions?

It is the single most trustworthy source because it is observed behavior, not an estimate, but it only shows demand you already touch. Pair it with Keyword Planner and Trends for demand you have not reached, and with GA4 to see whether ranked traffic becomes revenue. GSC tells you where you stand; the others tell you where to go and whether it paid.

What is public data useful for in SEO?

Sizing the market underneath the searches. County Business Patterns tells you how many businesses of a given type exist in a county; the American Community Survey describes households and incomes. That grounds targeting decisions (which metros, which segments) in something sturdier than keyword volume, which is modeled and rounds hard.

How often should I review SEO data?

Monthly for the working read (queries moving, pages gaining or losing, conversions by landing page) and quarterly for the strategic read (demand shifts in Trends, market changes, re-running the keyword loop). Daily checking creates noise-chasing; annual checking is how sites decline unnoticed.

What are the four layers of SEO data?

Query data, meaning what you actually appear for and where. Outcome data, meaning whether any of it produced leads or revenue. Demand data, meaning direction and seasonality. And market data, meaning the size of the opportunity underneath the searches. Most sites have one layer and argue about strategy without the other three.

What counts as outcome data?

Leads, calls, bookings, and revenue attributed back to their source. That means conversion events configured on real actions rather than pageviews, call tracking per source because service businesses convert by phone, and a way to tie a closed job back to the search that started it.

Why is query data called the observed truth?

Because unlike volume estimates it is a record of what actually happened: the searches your site really appeared for, the positions it really held, and the clicks it really received. Every other layer involves modelling or inference. This one is measurement, which is why disagreements about strategy usually end the moment someone opens it.

What does demand data add that query data cannot?

Direction and season. Query data tells you what happened on your site; demand data tells you which way a market is moving and when it peaks. That is what stops a business building content for a season that has already passed, or reading a normal seasonal dip as a ranking failure.

How do I join the data layers into a monthly read?

One pass, in order: which queries moved, what that did to traffic, what that traffic did for leads, and whether the demand picture changed. Each step conditions the next, so a rankings report with no outcome data and an outcome report with no query data both leave you guessing about cause.

Do I need a paid SEO tool suite?

Not to start, and many small businesses never do. The free layer covers query data, outcome data, and speed measurement. Paid suites earn their place for competitor visibility, backlink data, and rank tracking at scale, which matter more as the site and the competitive set grow.

Why keep a change log?

Because without one the data lies to you. Six months later nobody remembers whether traffic moved because of the content, the site migration, a Google update, or a seasonal swing. A dated line per change, taking a minute to write, is what turns a chart into evidence rather than a Rorschach test.

What is market data and why does it sit underneath the searches?

It is the count of potential customers and competitors in an area, from sources like census and business-registry data. It sets the ceiling any search strategy operates inside. Strong search volume in a market with too few buyers is a number that leads somewhere unprofitable, and only this layer reveals that before you spend.

Question we did not answer? Ask us directly and we will answer it here.

Written by
John Cravey
Founder

Founder of Frontend Horizon. Writes most of the long-form work on the FH blog.

Newer post
How to Find Keywords for SEO: A Working Process That Does Not Need Enterprise Tools
Older post
How to Grow a Small Business: A Stage-by-Stage Playbook for Owners
Keep reading

More from the blog

Search Console·12 min

Measuring AI Search Visibility in Google Search Console

Google added generative AI reporting to Search Console. Learn what the impressions and clicks mean, how to read them without over-reacting, and how to tie them to real business outcomes.

Search Console·10 min

Turn Google Search Console Into an AI Opportunity Finder

You are already ranking on page two for things you never wrote about. An AI agent's job is to find them before you waste effort elsewhere.

Analytics·4 min

Joining GA4 with Search Console: The Reporting View That Tells You What Actually Works

Search Console shows what queries clicked. GA4 shows what queries converted. The join shows what actually matters.