Inside the Perimeter and outside the Perimeter should never share one campaign budget
The two halves of this market have different product, different pricing and different buyers, and averaging them hides which one is working.
Intown, an occupier is weighing an office near the BeltLine against one in Midtown, and the conversation is about talent and walkability. Outside I-285, the same size requirement is a decision about parking ratio, commute from the north suburbs and rent.
Run them separately. Separate campaigns, separate budgets, separate landing pages. A blended cost per inquiry across both tells you nothing you can act on, because the two sides rarely perform alike.
Inside each half, split further by corridor rather than by a circle around your office. Cobb and Gwinnett are both outside the Perimeter and they do not compete for the same tenant.
Set geography by drive time your team can actually cover. Buying clicks in a county nobody will drive to on a Tuesday afternoon is buying inquiries your dispatcher cannot serve.
The renewal and disposition searches are worth more than "office space for lease"
Space searches are crowded and cheap to enter; assignment searches are quiet and belong to the people who sign fee agreements.
Someone typing a phrase about a lease renewal, a lease audit or how to sell a building is a principal with a decision in front of them. Someone typing a generic space phrase might be a broker, a student or a competitor.
Bid on the work you want: tenant representation, landlord representation, disposition, site selection, sale-leaseback. Pair each with a county or submarket rather than the metro name.
Space terms still have a place, but only when they are specific enough to imply a requirement, with a size range, an asset class and a submarket attached. Broad ones burn budget on the way to nothing.
Keep the two groups in separate campaigns so the cheap traffic cannot eat the expensive traffic's budget. Platforms will happily reallocate toward volume, and volume is exactly the wrong target here.
Intown queries pull people buying houses, and your negative list has to say so
The residential market in this metro is enormous, and it will quietly consume a commercial budget through submarket names.
Buckhead, Decatur, Brookhaven and Sandy Springs are all residential search terms first. Add real estate to any of them and the auction fills with home buyers, agents and rental shoppers.
Build the exclusion list before launch, not after the first invoice. Rental, apartment, homes, agent jobs, license, courses, salary, internship. Then read the search term report weekly for the first month and keep cutting.
Exclude geographically too. Counties you will not serve, and states you have no business paying for, unless you are deliberately targeting out-of-market corporate real estate teams, in which case that belongs in its own campaign with its own budget.
Expect the negative list to get long. On a brokerage account it usually ends up doing more for cost per tour than any bid adjustment.
Run the ads on the hours your brokers pick up, not evenly from midnight to midnight
An inquiry that arrives while every broker is on I-285 is worth a fraction of one that arrives when someone can answer.
Commercial inquiries are worth calling back within minutes, and a brokerage where the whole team is out touring buildings mid-afternoon cannot do that. Look at when your calls actually get answered, then weight spend toward those hours.
Weekends deserve a hard look. Some brokerages get useful owner-user activity on a Saturday morning. Others pay for clicks that go to voicemail until Monday. Check yours before you decide.
Pace across the month against your own activity, not the calendar. Budget that flattens itself evenly will underspend the weeks you have new availability to push.
If nobody covers after-hours calls, either fund coverage or shut off the hours. Paying for demand you cannot answer is the most expensive habit in the account.
Every corridor campaign needs its own page naming the county, the access and the buildings
Traffic bought for one submarket and dropped on a homepage converts like a cold call.
A tenant clicking an ad about Gwinnett industrial should land on a page about Gwinnett industrial, with what is available, what the I-285 and interstate access looks like, and who covers it.
Include the drive-time framing on the page itself. Occupiers comparing two counties are comparing commutes for their staff, and a page that does that math for them is doing sales work.
Match the form to the campaign. A land campaign and an office campaign need different fields, and a single generic contact form flattens both into an inquiry nobody can route.
Keep the pages current. A landing page listing buildings that leased months ago costs you the click twice, once in spend and once in credibility.
A lease decision outruns the thirty day window your ad platform reports on
The account will look like a failure for months if you judge it on the reporting period the platform hands you.
Occupier decisions run long. An inquiry that arrives in March can sign in the fall, and the platform's default view will never connect them.
Push the CRM stage back into the ad platform, or at minimum tag inquiries by campaign and review them by cohort. Ask what the March cohort produced by October, not what last week cost per lead.
Price the thing that matters. Work back from a signed fee agreement to a booked tour to a qualified inquiry, and let that arithmetic set the bid ceiling. Every firm's numbers are their own, so build the model from your own pipeline rather than a benchmark.
Never let anyone promise a lead volume for a fixed spend in a market this fragmented. The honest commitment is structure, exclusions, pacing and a scoreboard you can audit.
Questions we actually get
- How much should we budget to test paid search here?
- Enough to buy a readable sample in one part of the market rather than a trickle across all of it. Pick one side of the Perimeter and one or two asset classes, fund those properly, and leave the rest off. Spreading a small budget across Fulton, Cobb, DeKalb and Gwinnett at once produces four campaigns that all starve.
- Our brokers say the paid leads are junk. What is usually wrong?
- Almost always geography and exclusions. Broad metro targeting plus submarket names that read as residential fills the account with home buyers and job seekers. Second most common cause is routing: a real industrial inquiry that sat with an office broker for a day looks like a bad lead by the time anyone calls it.
- Should we advertise buildings we do not have listed?
- Advertise the requirement, not inventory you cannot deliver. Running ads against buildings you do not control produces inquiries you have to hand away, and it damages the landlord relationships you are trying to win. Bid on the assignment and on submarkets where you have something real to show.
- Can paid search reach a corporate real estate team that is out of state?
- Sometimes, but not through local search alone. Those teams often use national portals and brokers first. Paid works better as reinforcement: target the submarket phrases they will eventually use, keep budget on your own submarket pages, and accept that the first contact may come from a colleague's referral rather than a click.
- Can we measure paid spend all the way to a signed lease?
- You can get close if the CRM is disciplined. Tag every inquiry with its campaign, keep a stage for booked tour, and review by cohort instead of by month. What you cannot do is expect the ad platform's own conversion reporting to reflect a decision that took most of a year.
