Two businesses can ask for the same thing here and need opposite programmes. One wants the phone to ring from a six mile radius. The other sells nationally and happens to be headquartered downtown. Both will say they want SEO in Chicago, and a supplier who does not force the distinction in the first conversation will end up building the wrong thing for one of them.
Local SEO in Chicago means ranking in a dozen separate places at once
For anything with a local intent, the results a person sees are shaped heavily by where they are standing. A firm in the Loop does not appear in the map results for somebody searching from a north suburb, and no amount of on page work changes that. The metropolitan area behaves as a cluster of adjacent markets that happen to share a name, and the practical unit of planning is the neighbourhood or the suburb, not the city.
Ask any SEO candidate how they will measure position by location rather than as a single citywide number. Rank tracking configured from one point produces a chart that is either flattering or depressing and is wrong either way. If your customers travel to you, the question is which catchments you can realistically hold; if you travel to them, the question is which catchments justify the drive.
Decide whether you are buying proximity or national reach
National sellers based here are competing with companies that have older domains, more coverage and larger content operations, and their programme looks nothing like a local one. It leans on topic depth, on earning citations from trade publications, and on the parts of the site that convert rather than the parts that attract. Local visibility is close to irrelevant to it.
Local sellers have the opposite economics. A modest number of well structured pages, a properly maintained listing, review flow and consistent business information will beat a large content programme every time. Paying for national SEO tactics to solve a proximity problem is the most expensive mistake available in this market, and it is common because national work is more interesting to produce.
Multiple locations need real pages, not a template with the town swapped
Businesses with several sites across the metro nearly always arrive with a folder of near identical location pages, each one a paragraph of boilerplate with a name substituted. These rank briefly and then stop, and in a competitive category they drag the whole domain down with them.
A location page earns its place when it contains things only that location can say: the people who work there, hours that differ, parking and transit realities, the services actually offered at that address, photographs taken on site, and reviews from customers who went there. That is a content production problem with an operational dependency, so ask who inside your company will supply it before you agree a page count. An SEO plan that promises dozens of location pages without naming that person is promising boilerplate.
Reviews are a ranking input here, not a reputation afterthought
In a dense market with many comparable suppliers, the volume, recency and text of reviews influence both what the map shows and what a person clicks. Yet review generation usually sits outside the SEO contract entirely, which means the single most movable local signal is nobody's job.
Settle it explicitly. Who asks, at what point in the transaction, through which channel, and who responds publicly. Ask the agency what it intends to do about review text as a source of language for the site, because the words customers use in reviews are frequently better keyword research than any tool output.
The SEO firms easiest to find are the ones best at marketing themselves
Search for SEO services in this city and the results are dominated by firms marketing themselves, which creates a peculiar evaluation problem: the companies easiest to find are the ones best at finding clients, and those are not the same skill as serving them. A strong ranking for their own name proves they can win an unusually contested term with unlimited time and no client approvals. Your project has neither.
So judge on client evidence instead. Ask for two accounts in a similar competitive position, with the specific pages they built and the measurable change, and ask what they were unable to fix and why. Ask how many clients each account manager carries. A supplier that is growing fast is often the one with the thinnest staffing behind the pitch.
Count calls and visits, not sessions
Organic traffic totals are a poor proxy for the value an SEO programme creates where a large share of outcomes are phone calls, direction requests and walk ins that never touch a form. Before work starts, instrument the things that actually indicate business: tracked numbers, click to call events, direction taps, quote requests and a way to mark which of those turned into revenue.
Then agree what counts as progress in the first two quarters. Non brand organic enquiries is a fair headline measure. Total sessions is not, because it rises whenever you publish anything at all.
Getting comparable answers out of a crowded Chicago field
Write one page of context, your locations, your service area, who deploys site changes and what you have already tried, and send it to three firms. Ask each for the first quarter in execution order, named people with weekly hours, and the one thing they would stop doing on your current site. Keep the assets in your own name from day one: search console and analytics properties created under your company account, listing management logins held by you, and all research and drafts delivered in editable files. If you would rather set the requirement out once and collect answers in a single format, describe the work and request offers.
Where organic search is one strand of a larger plan, agencies running several channels answer a different brief, and when earned coverage rather than rankings is the real goal, media relations firms are the better comparison. The overview of SEO agencies sets out what the category usually includes.