How I turn a messy warranty/SLA dump into a clean SLA tracker

A vendor PDF that says “99.9% uptime,” a support portal screenshot with response times in three units, a Slack thread that promises “priority coverage somehow,” and a calendar note labeled “warranty??” are not an SLA system. They are how “what’s covered, how fast, and until when?” stays fuzzy until something breaks and nobody can point to one answer.

Here is the workflow I use when someone dumps a messy warranty or SLA pack on me and wants one clean SLA tracker: every commitment named, timed, coverage-bounded, and easy to skim — without me pretending to be their lawyer, CPA, or telling them how to renegotiate vendor deals.

Why SLA trackers go wrong

Most dumps I see fail in the same three places:

1. Metric fog — “fast response,” “business hours,” and “best effort” mixed with real targets like 4-hour P1 or 99.9% monthly uptime
2. Coverage drift — warranty windows, included seats, regions, or product SKUs that do not match what they actually bought
3. Clock silence — start/end dates, renewal, and what starts the response clock never written down, so the sheet looks current when it is not

If those stay fuzzy, the “are we covered?” line lies. So I lock the commitments from the dump before I polish the formatting.

Sort into three piles first

Before columns, I sort every SLA- or warranty-looking clause into:

• Clearly stated commitments — metric + target + scope (or a rule that maps to a measurable line in the dump)
• Clearly expired or superseded — marked ended, replaced by a newer MSA, or out of warranty with enough evidence in the dump
• Unknown or conflicting — missing target, “TBD,” dual clocks, or coverage that contradicts the order form

Unknowns do not get silently forced into a neat green “compliant” grid. They get a status and a question for the owner.

What I ask for before I start

1. The dump (MSA/SLA PDF, warranty certificate, support portal export, email thread, or mix)
2. What product / service / order the SLA is supposed to cover (or “use what’s in the dump only”)
3. As-of date for “active vs expired vs unknown” (today, month-end, custom)
4. Any incident or ticket samples if they care about response/restore clocks (optional but helpful)
5. What “done” means (SLA tracker, optional upcoming-expiry list, optional question list for TBD lines)

I am a freelancer admin / spreadsheet helper. I do not give legal or financial advice, and I am not a CPA — this is commitment hygiene so their warranty and SLA terms are easier to see and hand off, not contract interpretation, credit advice, collections coaching, or tax timing.

Build the sheet row by row

1. Inventory every SLA/warranty artifact — PDF section, portal page, email clause; record source and as-of.
2. Normalize commitment IDs — “P1 response,” “uptime,” “hardware warranty,” “RMA window” → one canonical label per row; keep aliases in notes.
3. Lock targets from the dump — never invent a % or hour target to make the sheet look complete; if two docs disagree, flag both.
4. Resolve clocks only when supported — map “business hours” only if the dump defines the timezone and calendar; else leave “rule stated — clock undefined.”
5. Tag status honestly — active / expired / superseded / unknown (vs as-of); no silent “assume still covered.”
6. Hand back — the SLA tracker (+ short list of questions); no “you should escalate the vendor harder” advice.

Columns that usually work

Column	What goes here
Commitment ID	Canonical label (aliases in Notes)
Metric type	Response time / restore time / uptime / coverage window / other (as stated)
Target	Numeric or stated target from the dump; else “unstated”
Scope / coverage	Product, seats, region, severity, or “unstated”
Clock / measurement window	How the clock runs when the dump says; else “undefined”
Start / end (or warranty through)	Calendar dates when known; else stated rule + “date missing”
Status	Active / expired / superseded / unknown (vs as-of)
Source	Which PDF / portal / email the line came from
Notes	Conflicting targets, TBD clocks, “needs owner confirm,” dual docs

Common traps I flag instead of guessing

• Treating “priority support” as a numbered response-time SLA when no hours exist in the dump
• Blending uptime % with response-time hours into one fake “health” score
• Extending a hardware warranty past the certificate end date because someone “thinks it auto-renewed”
• Marking everything green “met” when the dump never defines how incidents are measured
• Publishing a neat coverage summary while a big TBD / conflicting-MSA bucket sits unmentioned

Those become Notes or a short question list, not silent edits.

Soft CTA

If you want this done for you as a one-off: I turn a messy warranty/SLA dump into a clean SLA tracker sheet in about 24 hours.

Product: https://brittanybonds.gumroad.com/l/slasheet

Shop: https://brittanybonds.gumroad.com/

Catalog mirror: https://brittany-bonds.github.io/

Code ADMIN12 for $2 off.

No fake CPA or lawyer claims — spreadsheet cleanup so your response times, uptime targets, warranty windows, and coverage status are easier to track and hand off.