Vendor Continuity: The Line Item Smart Boards Budget For
Switching vendors for a lower bid can cost a Florida board more than it saves. Here is how to weigh tenure, transition cost, and when switching is...
Two references sit on the desk before a vote. Both are glowing. Both say some version of the same thing: good communication, showed up on time, would use them again. The board reads both, feels reassured, and moves on to the vote. Nobody on that board learned anything a marketing page could not have told them.
This is what most reference checking actually looks like, and it is close to useless. Not because the references are lying. Most of the time they are describing a genuinely good experience. The problem is the question being asked. "Were they good to work with" almost always gets a positive answer, because a board that had a genuinely bad experience with a vendor usually is not still on that vendor's reference list. The list itself is a filtered sample. Getting useful information out of it requires asking a different kind of question than the one most boards default to.
Reference checking is a method, not a formality to clear before a vote. Done well, it is one of the few tools a board has that a licence lookup and an insurance certificate cannot replace, because it is the only step that asks people who have actually worked with this contractor what happened when something did not go perfectly. This is a guide to running that call the way it produces a real answer.
Not all references carry the same weight, and the difference has nothing to do with how positive the reference is. It has to do with how wide a comparison set the person giving it actually has.
A single board member who oversaw one project with one contractor has exactly one data point. They can tell you whether that specific job went well. They generally cannot tell you whether this contractor is better or worse than the alternative, because they have nothing to compare it to. Their answer is honest and it is also close to meaningless as a signal, because a single good experience does not distinguish a genuinely strong vendor from an average one that happened to have a smooth run.
A community association manager who oversees several properties, and who has worked with a range of vendors across those properties over multiple years, is a categorically different kind of reference. That person has seen how this contractor performs against real alternatives, has likely seen them handle at least one job that did not go according to plan, and has a basis for saying whether this vendor is genuinely above average or just adequate. When a CAM reference says a contractor is one of the better ones they work with, that statement carries the weight of an actual comparison behind it. When a single board member says the same thing, it might just mean nothing went wrong on their one job.
This does not mean single-board references are worthless. It means they answer a narrower question, "did this go well," rather than the broader one a board actually needs answered, "is this vendor consistently good, including when things get difficult." Weight the two differently, and if the reference list a bidder hands over is entirely single-board contacts with no CAM among them, that itself is worth noting.
Every bidder will supply references, and every bidder has an incentive to hand over their best relationships. That is normal and expected. It also means the list itself is not neutral, and a board that only calls the names on the provided list is only hearing the story the vendor chose to tell.
A more useful approach layers two sources.
Asking a bidder directly for the name of a CAM who manages multiple properties, rather than accepting whatever three names come attached to the proposal, is a reasonable and normal request. A contractor with genuine, broad community-association experience should be able to produce one without hesitation. A contractor who struggles to name a multi-property reference may simply not have much depth of experience in this specific market, which is itself useful information, separate from whether the individual jobs they have done were good ones.
A reference call that only asks "were they good" produces a one-word answer and nothing else. The questions below are built to produce specifics the board can actually weigh, and they are ordered to build toward the one question that matters most: how the vendor behaved when something went wrong.
| Question | What it is actually testing |
|---|---|
| How long have you worked with this contractor, and on how many projects? | Establishes whether this reference has real depth or is describing a single job |
| What kind of work have they done for you, and how does it compare to what we are considering? | Confirms the reference's experience actually overlaps with your scope, not a different trade entirely |
| Walk me through a project that did not go exactly according to plan. What happened? | The single most revealing question on this list. Forces a specific answer instead of a general impression, and every real vendor relationship has at least one of these. |
| How did they communicate when there was a problem, compared to when everything was going fine? | Distinguishes a vendor who is only responsive when things are easy from one who stays reachable under strain |
| Did the final invoice match what was proposed, or were there change orders? If so, were they explained and approved before the work happened? | Tests financial discipline and whether the vendor treats a change order as a conversation or a surprise |
| Would you use them again for a similar scope of work? If not, why not? | The closing question. A hesitation here, even a slight one, is worth more than the words used to answer it. |
| Is there anything you wish you had known before you hired them? | Gives a reference permission to raise something without framing it as a complaint |
Ask every question of every reference, in the same order, even when the first few answers are glowing. Consistency across the call list is itself a data point. A vendor who gets the same specific praise from three unrelated references (for example, every reference independently mentions the same project manager staying available through a delay) is showing you a real, repeatable strength, not a lucky run.
An emailed reference request produces a written, careful, one-paragraph answer that has usually been reread before it was sent. It is the reference equivalent of a scripted testimonial, and it strips out almost everything useful described above: the pause, the tone shift, the detail that comes out only because a live conversation left room for a follow-up question. A phone call takes longer to arrange and rarely takes more than a few minutes once it happens. That few minutes is worth more than a page of written references, because it is the only format where a board can actually hear the hesitation in an answer and ask "what happened" on the spot instead of guessing at what was left out.
If a written response is genuinely the only option, at minimum send the change-order and problem-handling questions from the table above by name rather than a generic "please provide a reference," and follow up by phone on anything that reads as vague or overly brief.
The words a reference uses matter less than how they answer. A board running these calls should be listening for pattern as much as content.
A reference who answers the "would you use them again" question instantly and without qualification is telling you something different from one who pauses, then says yes. The pause itself is information. It usually means the honest answer is more complicated than a clean yes, and it is worth a gentle follow-up: "it sounds like there's more to that, what happened?"
If a reference cannot produce a single example of something not going perfectly, even when asked directly, that is not proof the vendor is flawless. Every vendor relationship of any real length includes at least one hiccup, a weather delay, a material substitution, a miscommunication. A reference who insists nothing has ever gone wrong is either describing a very short relationship or is not being fully candid, and either possibility is worth noting.
"They were great" is not a strong signal on its own. "They caught a drainage issue during the walk that wasn't in the original scope and flagged it before it became a change order" is a strong signal, because it describes a specific behavior the reference actually observed, not a general feeling.
A reference who brings up something unprompted, whether positive or negative, is usually giving you their genuinely top-of-mind reaction to this vendor. Pay closer attention to what comes up before you ask your first question than to the answers you had to prompt for.
| Weak reference signal | Strong reference signal |
|---|---|
| General praise with no specific example ("they were great, very professional") | A specific incident described in detail, including what the vendor did and why it worked |
| Single project, short relationship | Multiple projects over multiple years, ideally across more than one property |
| Claims nothing has ever gone wrong | Describes something that did not go according to plan and how it was handled |
| Answers quickly without engaging the question | Takes a moment to think through the answer, indicating they are actually recalling specifics |
| Only available for a very short call, deflects deeper questions | Willing to talk through the relationship in real detail, offers to be contacted again |
| Reference cannot describe how the vendor handled a change order or a delay | Reference can describe the exact process the vendor used to communicate a change before it happened |
Two references, both hand-picked by the bidder, is not enough of a sample to make a meaningful decision on. A reasonable target is three to five, drawn from a mix that avoids relying entirely on the vendor's own curated list.
A board pressed for meeting time understandably wants to shortcut this to a single call. Resist that. The entire value of reference checking comes from the comparison across multiple conversations, not from any one call in isolation. A single glowing reference and a full set of five with real texture to the answers are not the same level of due diligence, even if both technically satisfy "we checked references" on a meeting agenda.
| Mistake | Why it undermines the check |
|---|---|
| Only calling the names the vendor provided | The list is curated by definition; it will not surface a weak or problematic relationship |
| Asking only "were they good to work with" | Produces a near-automatic positive answer that tells the board nothing it could not have guessed |
| Treating a single reference as sufficient | One data point cannot distinguish a genuinely strong vendor from an average one with a lucky run |
| Not asking about a problem or delay directly | References rarely volunteer a negative example unprompted; the direct question is what surfaces it |
| Weighing a single-board reference the same as a multi-property CAM reference | The two answer fundamentally different questions and should not be scored identically |
| Skipping references entirely for a vendor the board has heard good things about informally | Informal reputation is not a substitute for a structured comparison across a real reference set |
Reference checking is one part of a larger vetting process, and it answers a question the other parts cannot. Confirming a contractor is properly licensed and insured establishes that they are legally permitted to do the work; it says nothing about whether they are the right fit for a community association specifically. What a Licence Check Does Not Tell You covers that gap in more depth and is a natural companion to this guide: verification tells you a company is allowed to bid, references tell you how they actually behave once they are working on occupied residential property.
Reference checking also overlaps with, but is distinct from, evaluating how a vendor communicates once a job is underway. That is its own standard, covered in Evaluating Vendor Communication, and it is worth reading alongside this guide because several of the reference-call questions above are really asking the reference to describe that same communication standard from their own experience.
A separate but related question is how much a vendor's tenure with your specific property is worth once they are already working with you, and what it costs to switch even when a new bid looks attractive on paper. Vendor Tenure and the Cost of Switching covers that side of the decision, for boards weighing whether to put an existing, known vendor back out to bid at all.
Once a vendor is selected and a job is underway, reference checking on the front end pairs naturally with a real close-out standard on the back end. The Crew Close-Out Walk covers what a board should be able to require at the end of a job, which is in many ways the next reference a future board will be calling about.
No. It is a normal and reasonable part of due diligence to ask a bidder directly for a reference at a specific property they have mentioned, even if it was not on their formal reference list. A vendor with genuinely strong relationships across their work should have no issue with this request.
Three to five is a reasonable request for most projects, and an established contractor working in the community-association space should be able to produce that range without difficulty. A vendor who resists providing more than one or two, or who seems reluctant to have the board reach a CAM reference directly, is worth a direct follow-up question about why.
Treat it as a reason to widen the sample rather than accept it at face value. Try adding a reference outside the vendor's own list, and consider whether the references provided share a common trait, such as all being short relationships or all being the same type of project, that would explain the lack of texture in the answers.
Not automatically. Look at what specifically drove the lukewarm response and whether it is relevant to your project. A reference describing a mismatch in project size or scope is less concerning than one describing a communication breakdown during a problem, since the latter is more likely to repeat regardless of project type.
As a general rule, yes, because a CAM managing multiple properties has a wider comparison set to draw the answer from. That does not make a board-member reference worthless. It simply answers a narrower question, whether one specific project went well, rather than whether the vendor is consistently strong across a range of work and conditions.
A reference from within the last year or two is generally the most reliable indicator of current performance, since staffing, ownership, and quality control can all shift over time. An older reference is still worth including if it is the only way to see how the vendor has performed over a longer relationship, but weight it alongside at least one more recent one.
Element Service Solutions holds itself to the same reference standard described in this guide and welcomes a board reaching out to any community we currently serve. If your board is weighing a vendor decision and wants a second opinion on the property itself before the vote, we offer a complimentary on-site inspection to walk the scope and help the board ask sharper questions of every bidder on the table.
Switching vendors for a lower bid can cost a Florida board more than it saves. Here is how to weigh tenure, transition cost, and when switching is...
"Responsive" is not a standard, it is an adjective. Here is how a Florida board writes vendor communication into the award instead of hoping for it.
Splitting one project across three trades does not split the work three ways. Learn what a board and property manager actually absorb when nobody...