Most competitor tracking produces a chart nobody acts on. Five domains, a hundred keywords, lines crossing each other, and a monthly meeting where somebody says "they're ahead of us on that one" and nothing changes.
The takeaway up front: competitor rank tracking is only useful when it answers one question — which specific keywords could realistically move, and what is the page that currently beats me doing differently? That means picking a small number of genuine search rivals, tracking a keyword set you have a claim to, and reading a gap number rather than a rank number.
Done that way it becomes the most efficient prioritisation tool in SEO, because it tells you where the ceiling is low. Done the usual way it is a wall decoration.
Your search competitors are not your business competitors
This is the first mistake and it quietly wastes the whole exercise. The companies your sales team loses deals to are not necessarily the domains occupying the results for your keywords. Search competitors are whoever holds the positions you want, and that group routinely includes:
- Publishers and review sites ranking above every vendor on commercial queries.
- Marketplaces and aggregators that own the transactional terms.
- Adjacent-category tools that solve a neighbouring problem and rank for shared informational terms.
- A much smaller company than yours with one exceptionally good page.
Track the domains that hold the positions, not the ones on your competitive battlecard. If a comparison site sits at position 1 for your money term, that is the entity you are actually competing with for that click, whatever your commercial rivalry looks like.
How to identify them properly
Three sources, in order of reliability.
1. Your own SERPs. Take your 20 highest-value keywords and record every domain appearing in the top 10 for each. Count the appearances. The domains that recur across many of your terms are your real search rivals, and this method costs nothing but attention.
2. Search Console query data. Pull the queries where you already earn impressions with a poor average position. These are terms Google associates with you but does not reward — which makes them the fairest head-to-head ground. Anything you have never appeared for is an aspiration, not a competition.
3. Category knowledge, used sparingly. Add one aspirational domain, clearly labelled. Useful as a benchmark, dangerous as a target — you will start copying a site with a different authority profile entirely.
Pick three to five and stop. Beyond five, the chart becomes unreadable and every competitor added dilutes the attention available for the ones that matter. Re-review the list quarterly, not weekly — swapping competitors mid-quarter destroys your ability to compare periods.
Choose the keyword set you actually contest
The tracked set determines whether the output is actionable. Build it from three buckets:
- Contested terms — you and at least one rival both rank in the top 20. This is the core of the set and where movement is achievable.
- Their terms — they rank top 10 and you rank 21 to 50. This is the opportunity bucket, and it should stay small; a term where you are nowhere is a content project, not a tracking line.
- Your terms — you hold top 5 and someone is climbing. This is the defence bucket, and it is the one people forget until they lose a position that was paying the bills.
Deliberately exclude branded terms of any kind. Your brand terms flatter you, theirs flatter them, and neither tells you anything about competitive strength. The one exception worth tracking is competitor brand terms if you run comparison content targeting them — but keep them in a separate group so they never contaminate the aggregate.
A workable size for a small site is 40 to 80 keywords. Enough to see a pattern, small enough that every line can be explained.
Set the tracking up so the comparison is fair
Comparison only means something when the conditions are identical for every domain. Three settings decide this:
Location. Set the same location for all tracked domains, and set it where your customers are. A national average is fine for a national business and useless for a local one, where the pack and the localised results differ street to street.
Device. Track mobile and desktop separately when you serve both. They are different result sets, and a rival can be well ahead on one and behind on the other — averaging the two hides exactly the difference you would act on.
Frequency. Daily checks for the contested and defence buckets, so you can tell a genuine change from a wobble. Check less often and you see the outcome of a movement, not its shape.
The mechanics — locations, devices, keyword grouping, baselines — are covered in the keyword rank tracking guide. The only addition for competitor work is that every setting must apply identically to every domain, including yours.
Read the gap, not the rank
Here are the four numbers worth reporting. All of them are more useful than a list of positions.
Share of top 10. Of your tracked keywords, what percentage of top-10 slots does each domain hold? This is the single best summary of competitive visibility, it survives individual keyword noise, and it is the number to put in front of a client or a manager.
Average gap per keyword. Your position minus theirs. Negative means you are ahead. Sorting by this column produces your work queue instantly, because the keywords with the smallest negative gap are the cheapest wins available to you.
Head-to-head win rate. On how many tracked terms do you outrank each rival? Blunt, easy to explain, and it moves slowly enough to be trustworthy.
Movement correlation. When you and a rival both move on the same day across many keywords, that is an algorithm update or a SERP layout change, not something either of you did. When only one domain moves, something changed on that site. This distinction is the whole discipline described in telling a real ranking change from noise, and running competitors alongside your own data is the cheapest way to make the call — a rival's positions are a free control group.
One thing to keep in view: the presence of SERP features changes what a position is worth. A page holding position 3 beneath a large feature block may receive fewer clicks than a position 5 on a clean results page. Track feature presence alongside position, and be careful about celebrating a rank improvement on a query whose result page has changed shape.
Turn the gap into a work queue
Filter your tracked set to keywords where the gap is 1 to 5 positions and you sit in the 5 to 20 range. That list is your month.
For each one, open both result pages and answer four questions about the page that beats you:
- Format match. Is it a different content type to yours — a comparison where you published a how-to, a tool page where you published an article? Format mismatch outranks almost every on-page factor and is the most common reason a good page loses.
- Coverage. What sub-questions does it answer that yours does not? Look at its headings, not its word count.
- Freshness. Has it been substantively updated recently, or does it merely carry a recent date? Substantive updates move things; date stamps do not.
- Internal support. How many of their own pages link to it, and how prominently? This is usually the cheapest gap to close on your side.
Then pick the three keywords with the smallest gaps and the clearest diagnosis, act on them, and watch the tracked positions for two to three weeks. Keywords sitting just off page one deserve special attention — the promotion mechanics for those are set out in moving striking-distance keywords onto page one.
Resist fixing everything at once. Change ten pages in a week and you cannot attribute the resulting movement to any of them — and attribution is the reason you are tracking at all.
A weekly ten-minute routine
- Scan the movement view for changes greater than three positions on tracked keywords.
- Check whether rivals moved on the same terms — if everyone moved, log it as a market event and stop.
- Note any defence-bucket keyword that lost a top-5 position and investigate that first.
- Re-sort the gap column and confirm this month's three targets are still the right ones.
- Record what you shipped this week, dated, so next month's movement has something to be attributed to.
That last step is the one most people skip, and it is what turns a tracking dashboard into evidence.
FAQ
How many competitors should I track? Three to five. Enough to distinguish a market-wide shift from a site-specific one, few enough that every line is explainable. Add a sixth only by removing one.
Should I track competitors on the same keywords as myself? Yes — that is the entire point. Comparing positions across different keyword sets produces a number with no meaning. One shared set, identical location and device, for every domain.
What if a competitor outranks me on everything? Then you are benchmarking, not competing, and you should also track a domain closer to your own size. Aspirational tracking tells you where the ceiling is; peer tracking tells you what to do on Monday.
Do I need daily checks? For contested and defence keywords, yes — daily data is what lets you separate a genuine change from ordinary fluctuation. Aspirational terms you are nowhere near can be checked less often.
Does outranking a competitor mean I get their traffic? Not proportionally. Click distribution depends heavily on the shape of the result page, including features, ads, and how many results appear above the fold. Treat position gains as a leading indicator and confirm with clicks and conversions.
Pick the rivals that actually hold your positions, track one shared keyword set under identical conditions, and work the smallest gaps first. Add your rivals alongside your own keywords and watch your competitors' positions daily with SBRanker.
