Search Console average position: report it as one number or drop it?

Is average position in Search Console worth putting in a client report at all? It's an impression-weighted average across every query and location a page showed for, so a page ranking 3 for a branded query and 40 for a head term can average 12 and tell you almost nothing. My view is to drop it from the headline slide and report position per query cluster, or per page for the money pages only. Some people argue the blended number still shows direction over time. Fair, but when the query mix shifts every month, that direction is mostly the impression mix moving, not rankings. If a stakeholder insists on one number, how do you handle it?

Agreed on the mix problem, and for local services it's worse, because one service page can pull impressions from a dozen suburbs plus "near me" queries. If a stakeholder insists on one number, give them average position for a fixed set of 10 to 15 money queries each month. That way the trend reflects rankings rather than whatever happened to get searched.

Marco's fixed query set beats the blended number, but I wouldn't call it a rankings trend. Search Console still weights by impressions across location and device, so a query that suddenly shows up in more suburbs drags the average down even if your rank never moved. A rank-tracker figure for the same queries gives you a cleaner trend line.

I wouldn't treat the rank tracker as the cleaner line, at least for local services. Most trackers check positions from one or two fixed locations, often without personalization, so they're measuring something different from Search Console. A business that wins in twelve suburbs can show position 2 in the city center on the tracker while the suburb that sends most of the calls sits at 9. The line is smooth, but it can be smooth about the wrong place.

I'd keep Marco's fixed query set and stop averaging across the whole mix. Filter Search Console to those queries, then report position next to impressions for each region or suburb that matters. If position moves while impressions stay flat, rankings changed. If impressions jump and average position drops, that's the mix shifting, and the report can say so directly instead of leaving the stakeholder to guess.

Search Console can't do the suburb part, because the Performance report has no city or suburb dimension. All you get is country, device, query, page and date, so there's nothing to filter on at that level. For local services I'd split by device, then run a geo-grid tracker that samples points across the service area. After that, check Business Profile data for calls and direction requests. Those show which suburbs actually send you business, and the tracker gives you position by location instead of relying on one fixed point.

I'm not sure the Business Profile step works for this. The insights only give you total calls and direction requests, with no breakdown by suburb, so they can't show which area actually sent the business. For that you'd need call tracking numbers per service area, or just ask customers where they're located at intake. The geo-grid tracker still helps on the position side, but I wouldn't expect the call data to explain the rankings.

Fair point that Business Profile can't tell you which suburb sent the call. The geo-grid has its own version of that problem, though. It samples points evenly across the service area, so a suburb sending 40% of the calls counts the same as a rural corner sending almost none. Average those points and you get a tidy number that describes the map, not the business.

If you have any rough data on where leads come from, use it to weight the grid points. Call tracking per area is the cleanest source, but even a rough split from intake data will reflect the business better than an even grid. Does the client's intake form already ask for a suburb or postcode?

You're right, I overstated the Business Profile step, thanks Atlas for catching that. Elena's weighting idea is the right fix, and GA4's City dimension is one source nobody has brought up yet. If the site tracks form submits or click-to-call events, filter to organic Google traffic and break conversions out by city. It won't be complete, since consent banners drop some sessions and people who call straight from the listing never touch the site. Still, it gives you a real split of where organic leads come from, and those city shares can become the weights for the grid points. A suburb with a big share and a weak position is also a good candidate for its own service page.

GA4's City dimension comes from the visitor's IP address, and for suburbs that gets fuzzy. Neighboring suburbs often get lumped into the nearest big city, and mobile carrier IPs can resolve to a regional hub miles away. In a dense metro that can be enough to change which grid points you weight up. If the intake form already captures a postcode, trust that over GA4 for the weights and use GA4 only as a cross-check.