Your analytics cannot see the channel producing most of your tours
Brokers send links by email and text, and every one of those visits arrives labeled as direct traffic.
Look at a typical brokerage analytics account and direct traffic dominates. Not because people are typing the URL, but because the site is a document that gets forwarded. The result is a program optimized for the small share of visits that came from search.
Tag the links brokers send. A simple campaign parameter appended to building page URLs, built into whatever tool they use, separates a forwarded listing from a cold search visit and shows which buildings are actually being circulated.
Instrument the phone. Call tracking with proper handling of your published number, and a capture of what page the caller was reading. Half your real inquiries never touch a form, and none of them appear in a conversion count.
Agree definitions before touching anything else. What counts as an inquiry, what counts as qualified, what counts as a tour. Without that, every later measurement is an argument.
Routing should follow the building's county, not the broker's seniority
The fastest qualified reply wins, and the fastest reply comes from the person who already knows the submarket.
A Tarrant County industrial requirement landing on a Dallas office desk gets a polite acknowledgment and a two day delay while it finds its way. The inquirer has already called somebody else by then.
Write a routing table: county and asset class in, named broker out, with a named backup for every row. Cover after hours and cover vacations. A rule that fails silently is worse than no rule.
Route on the building the inquiry references, not on where the inquirer is sitting. Someone in Denton asking about a Fort Worth warehouse is a Tarrant County requirement.
Measure time to first reply per route. Averages hide the problem. A single route that sits at a day and a half will drag the whole firm's conversion rate and will not be visible in a blended number.
Relocating occupiers in Plano are calling three brokerages, not one
Recent corporate arrivals have no established provider, so they shop, and they shop quickly.
Plano, Frisco, McKinney and Allen have absorbed years of corporate relocation. That produces a steady supply of occupiers with a requirement, a deadline and no relationship with anyone in North Texas.
They contact several firms the same day. Whoever replies first with something useful shapes how the others get judged. It is not a fair way to select a broker, and it is how the selection happens.
Cover the hours that match the callers. Inquiries come from Central time and from companies still headquartered on the west coast, which means late afternoon here is mid-morning for them.
Set a target for first response and staff it deliberately. An acknowledgment within minutes that names a real person and a real next step outperforms a considered reply that lands tomorrow.
Buyers here ask about the slab on the first call, so answer it in the first reply
The content of the first response is a conversion lever, and most firms treat it as a formality.
North Texas sits on expansive clay that shrinks and swells with moisture and moves foundations. Any experienced buyer looking at a forty year old building will raise it, along with roof age and storm history. Predictable questions deserve prepared answers.
Build a standard reply packet per building: survey, any geotechnical work you have the right to share, roof documentation, rent roll and the specification sheet. Send it with the first reply rather than promising it later.
Keep the language careful. Provide documents, do not characterize condition, and state that structural questions belong to a licensed engineer and insurance questions to the buyer's own insurer and counsel.
Template it so assembling that packet takes two minutes. Anything that takes twenty will not happen on a Friday afternoon, which is when a surprising share of inquiries arrive.
The tour confirmation email is the leak nobody looks at
A booked tour and an attended tour are different numbers, and most firms only track the first.
Distances here are real. A tour that looks reasonable on a calendar can involve forty minutes of freeway between two buildings, and an out-of-town buyer with two days in the Metroplex will drop the third stop.
The confirmation should carry everything needed to actually show up: full address, gate and parking instructions, the broker's mobile number, a calendar file, and an honest note about drive time between stops.
Send a reminder the morning of. Trivial to automate, and it recovers tours that would otherwise quietly evaporate.
Track attended against booked, by broker and by submarket. A gap in one column usually points at a scheduling habit rather than a marketing problem, and it is fixable in a week.
Change one big thing at a time and judge it on booked tours
No single metro sends a brokerage site enough traffic to settle small questions with a split test.
Honest arithmetic first. Detecting a small difference between two button colors requires a volume of inquiries a commercial brokerage will not see in a year. Running the test anyway produces a confident-looking result that is noise.
So test consequential changes, sequentially. Replace the form. Rewrite the first reply. Restructure the listing template. Give each one a window long enough to gather real outcomes and compare against the period before it.
Keep a dated change log alongside the numbers, and note market events. A quarter with an unusual number of large requirements will flatter whatever you happened to change that month.
Some things do not need a test at all. Routing an inquiry to the broker who covers that county, or answering within the hour, are not hypotheses. Fix them and move on to the things that are genuinely uncertain.
Questions we actually get
- We get very few form submissions. Is conversion work still worth it?
- Usually more so, because the leaks are proportionally more expensive. With low volume the work shifts away from split testing and toward instrumenting the phone and email paths, fixing routing, and improving the first reply. Those are process changes with visible effects, not statistical exercises.
- Can you run an A/B test on our listing pages?
- Sometimes, and often we would advise against it. Small changes need traffic a single-metro brokerage does not have, and an underpowered test gives a confident wrong answer. Testing a whole template against the previous one, over a long enough window and measured on tours, is the version that tells you something.
- What should we actually be measuring?
- Time to first reply by route, inquiries qualified, tours booked, tours attended, and agreements signed. Report them separately by county, because a blended figure across Dallas and Tarrant will hide whichever one is broken.
- Does call tracking hurt our local search rankings?
- Not if it is set up carefully. Keep your published business number consistent everywhere it appears publicly and use dynamic number insertion only for the tracked sessions. Done properly it is the only way to see the channel that produces a large share of your real inquiries.
- Should we gate offering memoranda behind a confidentiality agreement?
- That is a legal and client-instruction question more than a marketing one, and worth confirming with your own counsel and the owner. What we can say from the data side is that whatever the gate is, it needs to work on a phone in one sitting, because a document workflow that requires a desktop loses a meaningful share of serious principals.
