Every engagement starts with someone else’s backlog.
Vindex scores every story in it, flags the ones that will turn into defects and rework, and turns your review into a white-labeled report that carries your brand: evidence your client can act on, not an opinion they can argue with. Built for agencies and consultancies that review other people’s backlogs for a living.
The free evaluator scores any CSV in about 30 seconds, with no signup and nothing installed. Or see how a project works.
“Acceptance criteria are missing or untestable on 14 of the 20 highest-priority stories; recommended fixes are included per story.”
Stop being the bad guy about client requirements. Let the score say it.
Without evidence, telling a paying client their requirements aren’t buildable is your opinion against theirs, and you’re the one who looks difficult.
Every client, a new environment
New domain, new tooling, new conventions, new backlog, and no fast way to know what you're walking into until your team is already burning hours on it.
The bad-guy problem
Pushing back on a client's stories sounds like making excuses before the work starts. You need the gap named by something that isn't you: a neutral third party with no stake in the argument.
Fixed bids eat the rework
When a vague requirement turns into rework, the client's timeline slips but your margin takes the hit. The gap was visible on day one; there was just no tool to show it.
Same conversation, different ending: the score names the gap on day one, while it’s still cheap to fix, and it isn’t your opinion.
Here’s what it catches, before anyone builds it.
Four stories from the sample backlog in every agency workspace, scored by the same engine that scores your client’s stories. Findings come worst-first, in plain language.
- 12%NW-110“Support other things the customer might need later”
No concrete outcome and nothing a tester could verify. Recommended: Remove or replace with concrete, testable needs.
- 15%NW-106“Fix bugs”
Unbounded scope, impossible to estimate or call done. Recommended: List the specific defects with reproduction steps.
- 20%NW-102“Improve the system”
Vague scope with no measurable acceptance criteria. Recommended: State which system and which outcome is expected, then add measurable acceptance criteria.
- 85%NW-101“As a dispatcher I want to see live truck positions”
Clear actor, outcome, and testable behavior. Recommended: Consider splitting map rendering from data refresh.
Built for the firm doing the reviewing.
Agencies inherit a new backlog with every engagement, and the job of telling a client what isn’t buildable. Vindex turns that hard conversation into a deliverable.
The report does the pushing back
Telling a paying client their requirements aren't buildable sounds like an excuse before the work starts. Vindex names the gap with an objective, per-story score from a neutral engine, so it's evidence, not your opinion.
Your brand on every page
Every report is white-labeled: your logo, your colors, your voice. Clients never need a Vindex account and never see ours; they see a polished assessment from their agency.
A service line you can quote
Package a backlog audit you can price, deliver, and repeat with every engagement. Credits are prepaid and the exact cost shows before every run: a clean unit cost you can build a margin on.
Jira, Azure DevOps, Asana, and Monday exports, with no file-size or row limits; 100,000+ stories score unattended.
Re-score after the client revises and the report shows the improvement round over round.
Every story is de-identified before any AI model sees it, and Vindex never trains on your client's data.
Five steps, one deliverable.
A project moves through a clear lifecycle, draft → scoring → review → reporting → delivered, and you can re-run it any time the client revises their backlog.
Set up the project
Create a client workspace and a project for the engagement. Everything for that client stays separated.
Upload the backlog
Drop the client's CSV or Excel export. Vindex detects the sheet and suggests a column mapping automatically.
Score it
Confirm the exact credit cost, then let the job run unattended; every story gets a health score and plain-language findings.
Review and curate
Work the findings worst-first: read the reasoning, exclude junk rows, add analyst notes, curate recommendations.
Deliver the report
Build a white-labeled report and hand off the file. Re-score later to show improvement round over round.
A report that looks like it came from you.
The output isn’t a Vindex dashboard your client has to log into; it’s a polished, self-contained report you deliver under your own brand. Compose the sections, set the branding, and generate an immutable version you can hand off with confidence.
- Your logo, colors, and about-us copy on every page
- Executive summary with backlog-health scores at a glance
- Per-story findings in plain language, worst-first
- Curated recommendations your analysts approved
- Round-over-round improvement when you re-score
- Delivered as PDF, self-contained HTML, and an Excel data export
Re-score after the client revises their stories and the report shows the delta: progress a stakeholder can see, and a natural reason for the engagement to continue.
Pay for stories scored. Nothing else.
No per-seat fees, no subscriptions to babysit. Credits are prepaid, transparent, and refunded when a score fails, so the cost of a review is always something you can quote a client with a margin.
Every re-scoring round is included: score the backlog, the client fixes their stories, score it again: round two of the same stories costs 0 credits.
A story you've already scored in a project never bills again, and if a story can't be scored you're not charged. You only ever pay for new stories that get results.
Prepaid balance carries across projects and clients. A run only starts when your balance covers it, and the exact cost, split into previously scored and new stories, is shown before you confirm.
Making the case to a client? Put their team size and ticket volume into the savings calculator and walk them through what at-risk stories cost in rework: real numbers for the pitch, editable in front of them.
The playbook, in the open.
How agencies and consultancies run backlog audits, present hard findings, price the work, and protect fixed-bid margins.
What does a low-scoring story actually cost?
One story from vague to shipped defect: what a story that looks ready hides, what scoring flags in it, and why the agency absorbs the rework on a fixed bid. The guide to send a client who thinks thin requirements are survivable.
How do I audit a backlog when the project is already slipping?
The mid-flight audit, where you are the vendor and the delay is partly yours: score what remains rather than re-litigating delivered sprints, split the slip into named causes, and turn the findings into a re-baselined date.
How do I estimate work from a backlog I didn't write?
Split an inherited backlog three ways before you price it: what you can size, what needs one question answered, and what cannot be sized as written. How to quote each bucket and present the split as rigour rather than hedging.
The evidence behind the audit.
Case studies and field guides on what bad requirements cost, the same evidence base your reports draw on.
How a Healthcare Platform Cut Rework 68% by Shifting Quality Left
A healthcare technology company was losing 340 engineering hours a month to unclear requirements. After scoring story health before code, they cut rework 68%, recovered $564K a year, and caught HIPAA gaps in refinement instead of UAT.
Bad Scores Cost Real Money: A $1.2B Automotive Recall Study
Inside a pilot where an automotive manufacturer ran 400,000 Jira tickets through Vindex. The lowest-scoring requirements traced to $1.2 billion in software-related recall costs. The highest-scoring traced to zero.
Stop Bad Requirements: Why Small Dev Teams Need Vindex
A story looks ready on Friday and blocks a developer on Monday. Here is how small teams catch vague, oversized, and untestable work before it burns their scarcest resource: developer time.
Make requirement review your next service line.
We’ll switch on agency features for your firm, invite your first admin, and add starter credits so you can score a real client backlog and deliver a report before any payment conversation.