An account targeting Missouri quietly turns off half the metro
Geographic settings written at the state level either drop the Metro East entirely or buy it by accident, and both happen constantly.
The Illinois side of the river is part of this market. Belleville and Edwardsville sit in St. Clair County and Madison County, an easy drive from downtown, and occupiers there are shopping the same corridors your brokers work. A Missouri only setting hides all of it.
The reverse error is broader. Targeting the whole metro area as one blob mixes the City of St. Louis, St. Louis County, St. Charles County and two Illinois counties into a single auction, then reports one blended cost per lead that tells you nothing about any of them.
Build geography around where the building is, and set the targeting option that includes people searching about the area rather than only people standing in it. A corporate real estate director in another state can be your best searcher, and presence based targeting removes him.
Then separate the reporting even where you keep budgets combined. Municipality level campaigns are often too thin to fund individually, but geographic reporting costs nothing and shows you which places are actually producing conversations.
Say the square footage floor in the ad, and let the wrong click go elsewhere
Copy that pre-qualifies is cheaper than an exclusion list you have to maintain forever.
Ad copy has a job before it has a voice. Naming a size range, an asset class and a jurisdiction turns the ad into a filter, and the clicks you lose were the ones you did not want to pay for.
Commercial phrasing in this market pulls a predictable amount of residential traffic. Anything touching brick or masonry attracts homeowners hunting a tuckpointing bid. Warehouse and property phrases attract job seekers. Investment phrasing attracts students and course buyers.
Read the search terms report weekly for the first stretch, then monthly. Add exclusions in groups rather than one at a time: residential intent, employment intent, education intent, do it yourself repair intent. Most accounts need a few hundred over a year and then settle.
Keep the exclusions attached to real evidence. Blocking terms because they look unserious, without checking whether they ever produced a conversation, is how accounts get quietly starved.
The owner user shopping a small brick building is a campaign nobody is running
Small businesses buying their own building search like consumers and are almost never sold to properly.
A contractor, a clinic or a small manufacturer outgrowing a lease starts by asking whether buying beats renting. The searches look nothing like a tenant rep's and the competition for them is usually thinner.
The inventory that fits is often an older brick building in a city neighborhood or an inner ring suburb, which means the ad has to answer condition questions early. Age, roof, masonry, loading, parking, and whether the use is permitted where it sits.
Financing questions belong with a lender and legal questions belong with counsel, so the copy and landing page should say so rather than implying your firm can answer them. Point the reader at the right professional and keep the offer to what you actually do.
Fund the campaign small and separately. Volume is limited, so a shared budget lets the busier tenant campaigns eat it inside a week and you never learn whether it works.
Feed the bidding a real conversion, or it will buy you flyer downloads
Automated bidding optimizes toward whatever you told it to count, with no opinion about whether the thing is valuable.
Most brokerage accounts count everything: newsletter signups, PDF downloads, map clicks, any form on the site. The algorithm then finds the cheapest of those, which is almost always the download, and spends the month buying them.
Define one primary conversion that means a real inquiry, and demote the rest to secondary so they are visible but not steering the bid. A phone call over a duration threshold usually belongs in the primary set, because a large share of commercial inquiries arrive by phone.
Where your systems allow it, push the later stage back into the platform so bidding sees which inquiries became conversations. If that is not workable, a simple weekly count of qualified inquiries by campaign, kept by hand, beats optimizing toward the wrong signal.
Check the tracking after every site change. A form migration that quietly stops recording conversions will not announce itself, and the bidding will drift for weeks before anyone notices the lead count fell.
Check whether your tour rate moves with the first hard freeze before you cut winter spend
Seasonality is real here, but the version in your head is probably not the version in your data.
Winters in this region are cold enough to change behavior. Touring a vacant unheated brick warehouse in January is a different experience than touring it in April, and freeze-thaw damage can take a building off the market for a while.
None of that justifies rewriting the account in December. Pull two or three winters of your own data, compare inquiry to tour rates by month, and decide from that rather than from a story someone tells at a sales meeting.
Pacing rules should be slow. A quiet week during a cold snap is not a signal, and accounts get destroyed by reactive edits far more often than by neglect. Set a review cadence and hold to it.
If the data does show a real seasonal dip, the sensible response is usually to shift emphasis rather than cut. Owner side and landlord representation messaging tends to have its own rhythm, and winter is often when those conversations start.
A more expensive Clayton inquiry can still be your cheapest closed deal
Comparing cost per lead across geographies only works if the deals behind them are the same size, and they never are.
Cost per lead is the number everyone reports and the number that misleads most. An office requirement in Clayton and a small bay inquiry in Florissant cost different amounts to acquire and are worth different amounts to the firm.
Work backward instead. Estimate what a signed assignment is worth by geography and asset class, apply your own inquiry to tour and tour to signed rates, and derive what you can afford to pay for an inquiry in each. Use your numbers, not benchmarks from somewhere else.
Judge the account on booked tours and signed agreements, not on form fills. In a market this size the sample is small, so evaluate over quarters and resist moving budget on a single month's result.
Write the definitions down before you start. What counts as a qualified inquiry, what counts as a tour, who records it. Accounts usually fail on bookkeeping rather than bidding.
Questions we actually get
- Should tenant campaigns and landlord campaigns share a budget?
- Generally not. They are different buyers with different search language, different economics and different sales cycles, and combining them means the cheaper clicks win the budget by default. Two funded lines, even small ones, keep the assignments you most want to earn from being crowded out. Keep the reporting separate as well, because a blended cost per lead across both is not a number you can act on.
- How much should a brokerage spend to start?
- Enough for one geography and one buyer type to gather readable data, which is a question about your market and your deal sizes rather than a fixed figure. Spreading a small budget across the city, the county, St. Charles County and the Metro East usually produces four inconclusive campaigns. Start narrow where you have the strongest inventory or the strongest track record, prove the economics, then widen.
- Do we need a separate landing page for each campaign?
- You need a page that answers the promise in the ad. If the ad names a size range and a jurisdiction, the page should open with matching inventory rather than a general homepage. Whether that is a purpose built page or a filtered availability view matters less than relevance and speed. One good page per campaign beats twenty thin ones nobody maintains.
- Is LinkedIn worth running alongside search?
- It reaches a different moment. Search captures someone already looking, while professional network advertising reaches people before they are. For brokerages chasing landlord assignments it can be worth testing, but it is a slower channel with a different measurement rhythm, and it should not be funded out of a search budget that is currently producing conversations.
- Our phone rings more than our form fills. How do we measure that?
- Use call tracking with a duration threshold so short misdials do not count, and keep separate numbers by campaign so calls attribute correctly. Then log what happened on the call, because a call is not a qualified inquiry until someone says it is. Without that step you end up optimizing toward ring volume rather than toward conversations that lead to tours.
