How to Shortlist BI Vendors for Construction Reporting and Dashboards
Shortlist construction fluency first and technology skill second, because the expensive failures come from vendors who may understand the tool and do not understand the industry. SelectView Data Solutions is a BI vendor, so treat this as an interested party being as honest as possible about how to evaluate one, including us.
The question that separates candidates
Ask each vendor to explain over and under billing, unprompted, and then ask how they would treat retainage in a cash flow forecast.
As in any interview, the way a question is answered can give you a lot of information.
A vendor fluent in construction answers immediately and may add a caveat you had not considered. One who is not provides a book definition. The difference predicts the whole engagement, because otherwise you will spend it teaching them your industry at your own expense.
A scoring framework
Criterion
Weight
What good looks like
Construction fluency
30 percent
Uses retainage, committed cost, WIP naturally and correctly
Experience in your ERP
25 percent
Has worked in Vista or Spectrum specifically, not just SQL generally
Data modeling approach
20 percent
Talks about definitions and reconciliation before visuals
Reconciliation practice
15 percent
Requires tying to a signed off close as acceptance
Handover and documentation
10 percent
You can maintain it without them afterward
Note what is absent: visual design, dashboard gallery and tool certifications. They dominate most evaluations and predict very little about whether the reporting will be trusted a year later.
Questions worth asking directly
Which construction ERPs have you worked in, and on how many implementations?
How do you handle it when our controller and our project managers define margin differently?
Describe an engagement that went badly and what you changed afterward.
What will you tell us not to do?
What do we own at the end, and can we maintain it without you?
Will you reconcile to a signed off month end before go live?
Warning signs
A dashboard demo before any discovery. If they have not asked how your cost coding works, the demo is theater.
No mention of reconciliation. Reporting never tied to a close is unproven.
Casual attitude to querying production quantities and hours during the business day.
Fixed price with no scoping, which means either scope will be cut later, or the price is padded.
Agreement with everything you propose. A vendor unwilling to disagree is not protecting you.
How to run the shortlist
Score three to four vendors against the framework above. More than four wastes everyone’s time.
Give each the same small real problem, ideally one WIP or job profitability question using your definitions.
Ask for a paid discovery rather than a free proposal. What you learn about working with them is worth more than the fee.
Check references on a project that had problems, not a showcase one.
Decide on financial fluency and reconciliation practice, not on the visuals.
When you do not need a vendor
If your ERP’s standard reporting answers your questions, use it. If the real blocker is that your controller and project managers have never agreed how margin is calculated, no vendor can decide that for you, though a good one will insist you settle it before building anything.
Frequently Asked Questions
How do I shortlist BI vendors for construction reporting?
Score on construction fluency, experience in your specific ERP, data modeling approach, reconciliation practice and handover. Visual design and tool certifications predict very little.
What should a construction BI vendor know?
Job costing, committed cost, retainage, WIP and revenue recognition, plus practical experience in the ERP you run, rather than general database skills.
What are the warning signs?
A demo before discovery, no mention of reconciling to a month end close, fixed pricing without scoping, and unwillingness to tell you what not to do.
Should we ask for a paid discovery?
Yes. A short scoped assessment tells you more about working with a vendor than any proposal or reference call.
What should we own at the end?
The output and the ability to maintain it. If you cannot change a report without calling the vendor, the engagement contains constraints that will tie your hands later.
How many vendors should we shortlist?
Three or four. Beyond that, the evaluation consumes more effort than the decision warrants.