A tenant rep carrying a dozen retail clients does not need a dozen processes. What works is one method applied to every client, criteria set per client from what that tenant's own stores respond to, and a workspace per client that nothing crosses. Set up that way, taking on a new assignment costs a conversation instead of a new spreadsheet.
That split is the whole job. Standardize too little and every client gets an ad-hoc answer you cannot compare to last month's. Standardize too much and you are scoring a fitness concept on drive-thru logic.
Why this is harder than the same work done in-house
An operator's real estate director sets the criteria once. They are picking sites for one brand, in one format, against one set of unit economics. The bar gets set, and it holds for years.
A tenant rep resets that on every assignment. The corner, the co-tenants, the rent, and the parking count are the same facts. What each client wants out of them is not.
Three things make the broker version distinct:
The work is an argument, not a decision. The operator's analysis exists to make a call internally. The broker's analysis exists to win an assignment, justify a tour, and hold up when the client's CFO asks where a number came from. Same math, different purpose, and the second one has a much lower tolerance for anything the client cannot follow.
The volume is per client, not per site. Most brokers we work with are not evaluating more sites than an operator would. They are evaluating them for more people, each with a different bar, and each expecting the turnaround of someone with only one client.
Separation is not optional. Two clients in the same retail category is normal in tenant rep. Their working files touching each other is not.
The layer you are allowed to standardize
Standardize the method. That means the same data sources with the same vintage, the same way you draw a trade area, the same scoring mechanic, and the same report format coming out the other end. Anybody in your shop should be able to open a package for any client and know where to look.
Tidiness is not the payoff. What you get is that a shortlist you built in March reads the same way as one you build today, for a different tenant, so you can actually say this corner is stronger for this client than the last three you showed them. Without a shared method, you are comparing prose.
Do not standardize the criteria. This is the mistake that produces the most confident wrong answers. The site selection criteria belong to the client, and they should come from that client's own store performance rather than from a house template. Multi-brand platforms hit the same wall from the other direction: private equity operators running several portfolio brands have to keep one standard across brands without flattening what makes each one work. A tenant rep is running that problem with clients instead of portfolio companies, and with less control over any of them.
Set the criteria in the first client conversation
The criteria conversation belongs at the start of an assignment, before the first tour, not after the client rejects three sites you thought were good.
Five questions get you most of the way there:
- Which of your current stores would you open again tomorrow, and which would you not? The gap between those two lists is the criteria, stated in the only terms that matter.
- What kills a site outright? Ceiling height, drive-thru stacking, a grease trap, a landlord who will not do a co-tenancy clause. Binary disqualifiers save more time than any weighting does.
- What is the rent number above which this stops working? Most clients have one. Many have never written it down.
- Who is the customer, and when do they show up? A concept that lives on the lunch hour needs daytime population. One that lives on the evening commute needs a different corner on the same street.
- Who has to approve this internally? A recommendation going to a founder is a different document than one going to a real estate committee.
Write the answers down and keep them with the client's workspace. On the next assignment for the same client you start from a bar rather than a blank page, and the client notices that you remembered.
Re-assembly is what actually costs you the week
The analysis is rarely the slow part. Rebuilding the package is.
The standard broker stack is real: CoStar for the listing and the comps, Placer.ai for visitation, Esri for demographics, and a spreadsheet holding the parts that do not fit anywhere else. Each of those tools is good at its job. The cost sits in the handoffs between them, which is the same problem operators hit with a fragmented site selection stack, multiplied by the number of clients you carry.
Then the client adds a corner. Or drops one. Or asks what the fifth option looks like next to the second. Every one of those restarts the chain, and the shortlist moves most in the days right after a tour, which is exactly when the client is paying attention.
The version brokers keep asking for is a link instead of an attachment. A static PDF is a snapshot of a shortlist that has already changed. A live view is the shortlist. It also removes the worst version of this problem, which is a client sitting in a meeting quoting a number from a deck you replaced two weeks ago.
Keeping two clients in the same category apart
This one is not a preference. NAR's Code of Ethics, Article 1, obligates a REALTOR representing a buyer, seller, landlord, or tenant to protect and promote that client's interests. Standard of Practice 1-9 goes further: the duty to preserve confidential information provided by a client continues after the agency relationship ends, and it bars revealing that information, using it against the client, or using it for the broker's own advantage (National Association of REALTORS, 2026 Code of Ethics).
In practice the failure is never dramatic. It is a spreadsheet with three tabs, one per client, sent to a fourth party by accident. Or one client's uploaded store list becoming the comparison set for a competitor's search because it happened to be the file already open.
A workspace per client fixes it structurally rather than through discipline. Shortlists, notes, uploaded store data, and pipeline live in one client's space and nowhere else. Inside a multi-producer shop the same rule applies between reps, because a pipeline visible to the desk next to you is a pipeline that has left your control.
What the client is actually buying
A client about to commit to a ten-year lease is not buying your rating. They are buying the reasoning underneath it.
There is a trust problem in this industry, and it cuts in the broker's favor if you handle it right. In a survey of 255 commercial real estate professionals fielded in late March and early April 2026, 66 percent said they use AI weekly or daily, while only 5 percent trusted it enough to inform an actual deal decision. The top barriers were not knowing which tools to use, at 34 percent, and doubts about accuracy, at 32 percent (First American Data & Analytics and DealGround, May 2026).
Your client's team carries the same doubts, so a recommendation that opens up beats one that does not. Show which variables moved it, where the data came from, what the trade area covers, and which comparable sites informed it. The mechanics of defending a site forecast are the same whether the room is your client's committee or your own pitch, and a score nobody can trace is the one that falls apart under the first hard question.
Market conditions raise the cost of getting this wrong. CBRE put US retail availability at 4.9 percent in the second quarter of 2026, with average asking rent at $24.79 per square foot, up 2.4 percent year over year (CBRE, July 2026). JLL counted 10.2 million square feet of net absorption in the same quarter, its second strongest in two years, against construction that stays historically constrained outside a handful of Sun Belt metros (JLL, August 2026). Tight space and thin new supply mean fewer options per client and less margin for a shortlist built on a hunch. At the same time Coresight Research projects 7,900 US store closures and 5,500 openings for full-year 2026 (Coresight Research, August 2026), which is a lot of second-generation space coming back to the market on somebody's phone call. The rep who can say by Friday which two of their eight clients that box actually suits is the one who gets the next call.
Where GrowthFactor fits
We built the broker side of the platform around this exact split: a workspace per client, criteria set per client, one method underneath all of them, and a report that carries the client's branding rather than ours. The broker page walks through what that looks like on a real shortlist.
None of that replaces the part of the job you are actually paid for. You know the landlord, the market, and which deal can genuinely be done. That is what buys the week back: you spend it on the landlord and the deal instead of on rebuilding the same package for the fourth time. If you want the wider view of what a broker does and where analytics fit in it, our guide to what a CRE broker does covers the ground around this piece.
Frequently Asked Questions about tenant rep site analysis
How do tenant rep brokers analyze sites for multiple clients at once?
They separate what must stay the same from what must not. One method carries across every client: the same data sources, the same trade area definition, the same scoring mechanic, the same report format. The criteria change per client, because a good corner for a drive-thru coffee brand is not a good corner for a gym. The working files stay in a separate space per client. Skip the first part and you end up with recommendations you cannot compare from one assignment to the next.
Should a broker use the same site criteria for every retail client?
No. The method should be identical and the criteria should not. A drive-thru concept lives on morning commute volume and turn-in geometry. A neighborhood gym is a destination trip, so pass-by traffic is close to noise for it. Value apparel cares more about the co-tenants and the parking count than the corner itself. One weighting applied to all three gives you numbers that agree with each other and disagree with what each tenant needs.
How do brokers keep two competing clients' site data separate?
With a separate workspace per client and a habit of never moving files between them. NAR's Standard of Practice 1-9 holds that the duty to preserve a client's confidential information continues after the agency relationship ends, and bars using it to that client's disadvantage or for the broker's own advantage. The practical version is that shortlists, notes, uploaded store data, and pipeline live in one client's space only, and a second client in the same category gets its own from day one.
What goes into a defensible site recommendation for a retail tenant?
The inputs, not just the answer. A client about to sign a ten-year lease wants to see which variables moved the recommendation, where the data came from, what the trade area covers, and what the comparable sites look like. A single rating with no visible math is the version that falls apart in the first hard question from a CFO. Brokers who show the inputs win the second meeting more often than brokers who show a conclusion.
How does GrowthFactor compare to Placer.ai for a tenant rep broker?
They sit at different points in the job. Placer.ai is a foot-traffic data provider and is strongest as a measure of visitation to places, which a broker can use across several clients. GrowthFactor is a site selection and deal platform, so it takes data like that as one input and produces a scored, explainable recommendation per client, with the variables and weights visible and a separate workspace for each assignment. Brokers often want both: a data layer underneath, and a decision layer that turns it into something a client can read.