Article

Do messy records still count?

By TrustPast

Published October 26, 2026 · Updated October 31, 2026

Every owner who looks at their own back office sees the gaps first. Work orders with half the boxes blank. Estimates that only exist as an attachment in somebody’s sent folder. A switch of software in 2019 that left six years of jobs in a system nobody logs into. So the question comes before any other: do records like these still count, or does a buyer only want the tidy kind? This page settles it, using micro1’s own words where it has them and labelling our own advice where it doesn’t.

Our cornerstone page on what data buyers want from a trade business sets out what is wanted, trade by trade. Here we take one worry and stay with it. The worry is that your records are too scattered, too incomplete and too inconsistent to be worth anything. In most companies, as our own reading, they are messy in exactly the way that holds value.

The short answer

Yes, messy records can count. micro1 doesn’t ask for clean files anywhere on its pages. Its October announcement says: “Real business operations involve incomplete information, competing priorities and exceptions that require judgment.” As our own reading, that is the opposite of what owners fear. Its records list asks for “data that reflects how work is actually performed within an organization.”

Two things keep that answer honest. micro1 does name quality as a factor in the money, and it reviews what you send “for completeness.” So, as our own reading, mess costs less than missing work. A work order with a blank field still shows a job. A year that was never kept shows nothing.

And micro1 is silent on some of the things owners ask most. It says nothing on its pages about handwriting, scans, duplicate copies or file formats. Where it is silent, we say so below, and the advice that follows is ours.

Apply & License Your Data

What micro1 says, word for word

We fetched four of micro1’s public pages on 26 October 2026. Here is every line that bears on messy records.

What it looks for. “We are looking for data that reflects how work is actually performed within an organization.” The list that follows names “standard operating procedures (SOPs), process documentation, internal knowledge bases, workflows, templates, operational frameworks, and other business process artifacts.” Its eligibility answer adds “internal communications, SOPs, project histories, CRM records, support tickets, and other business documentation.”

What it values most. “The most valuable datasets capture real operational expertise, decision-making processes, and repeatable workflows”.

How much it needs. “There’s no fixed minimum number of files, documents, or terabytes.” Then: “We evaluate the historical depth, breadth, quality, and relevance of the available data. Companies with several years of records across multiple business systems are generally a stronger fit.”

What it checks on delivery. Its third step reads: “You securely transfer your data through our APIs or other approved methods. We review the data for completeness.”

Whether you clean it first. Asked “Does my company need to anonymize its data before transferring it?”, micro1 answers: “No, micro1 has processes for de-identifying and transforming sensitive information.”

What sets the money. “Factors considered include dataset size and volume, workflow complexity, data quality, domain expertise represented, and the uniqueness of the operational knowledge contained within the dataset.” Its payout page says “Final payout is determined based on data sources, volume, quality, access terms, and due diligence.” On its referral page, company compensation is “$100K-$2M+ for approved data packages” (indicative, not an offer).

Why the mess matters. The announcement says micro1 uses de-identified company data to build environments that reflect real business operations. It says those environments need to “preserve the context behind how businesses work.”

Read together, as our own reading, the lines point one way. Scattered across systems is a plus, not a minus. Incomplete information is part of what is being studied. Quality still counts, and so does having the whole job on file.

Messy is not the same as low quality

Owners use one word for two different things. It helps to pull them apart. This split is our own, not micro1’s.

Messy is how real work looks on paper. Abbreviations. Blank fields. Notes typed with one thumb in a crawlspace. An estimate revised four times by email instead of in the software. Two systems that each hold half the story.

Low quality is when the work itself didn’t get written down. A ticket that says “fixed” and nothing else, on every job, for ten years. A folder of invoices with no estimates, notes or change orders behind them. A record that can’t be tied to any job at all.

The first kind still shows how the work was done, once you put the pieces side by side. The second kind shows only that something was billed. micro1 asks for how work is actually performed. A messy file can answer that. A tidy file that only holds totals can’t.

So when you look at your records, don’t ask whether they’re neat. Ask whether a stranger could follow one job from the first call to the last visit. If the answer is yes, even across three systems, the mess matters less than you fear.

Half-filled work orders

Take a commercial plumbing contractor with 40 people. Its work orders are the backbone of the job file. About a third have the cause field empty. Many have the parts typed into the notes instead of the parts table. Some techs write paragraphs. Others write four words.

Here is what a blank field doesn’t erase. The complaint is there: no hot water on the third floor. The time on site is there. The parts are there, even if they’re in the wrong box. The return visit two days later is there, with a note that the mixing valve was the real cause, not the heater. That chain is a repair being diagnosed, including the first guess that was wrong.

As our own reading, the wrong first guess is worth more than a clean form. It shows a judgment call, a check and a correction. That is the kind of exception micro1’s announcement describes. A perfectly filled form that says “replaced valve” teaches less.

Our advice: don’t go back and fill in the blanks. Blank fields do weigh on quality, as our page on what decides the price notes. Filling them years later doesn’t fix that; it swaps a gap for a guess. Leave the work orders as they are, and tell micro1 which fields are often empty. Our page on job files and work orders covers what a job file holds.

Estimates that live in email threads

Many companies price jobs in the software and then negotiate them somewhere else. The estimate goes out as a PDF. The customer replies with questions. The estimator answers, cuts a line, adds an alternate, and resends. The final number lives in the fourth reply, not in the system.

To an owner, that looks like a filing failure. As our own reading, it is the most useful part of the record. The thread shows how an estimate became a job. It shows which line the customer pushed back on, what the estimator gave up, and what they held. When the job hits a surprise, the same thread often carries the change order and how it was settled.

The software alone would show only the first estimate and the final invoice. The thread holds the reasoning between them. That is what micro1’s phrase “decision-making processes” covers, as we read it.

Our advice: find out where those threads sit. An estimator’s mailbox, a shared quotes inbox, an old mail server. Count the mailboxes and the years they cover. Our pages on estimates and quotes and email archives go further. If your mail is in Microsoft 365, our guide to exporting Microsoft 365 mail shows who can run the export. For Gmail, see the Google Workspace export guide.

Job notes in a tech’s phone

Some of the best notes in a service company never reach the office. A tech photographs the data plate. They text the shop a picture of a cracked heat exchanger. They keep a running note on their own phone: which units on which building trip in July, which property manager wants a call before anyone goes on the roof.

micro1’s pages don’t mention phones, texts or photos. They list “internal communications” among the records a company may hold. As our own reading, a text thread between a tech and the dispatcher about a job is an internal communication. A note on a personal phone is less clear. It may belong to the person, not the company.

Our advice, labelled as ours: start with what the company controls. Texts sent through a company line, photos attached to the job in your field software, notes typed into the work order app. Don’t ask staff to hand over their personal phones. If a large share of your know-how lives on devices you don’t own, say so on the first call. Our page on employee personal data covers the staff side. If those notes are written in Spanish or another language, our page on records not in English shows how to describe them.

If you use ServiceTitan, many of those photos and notes may already sit on the job record. Our ServiceTitan export guide shows how to count them.

The software switch in 2019

Here is a common one. A mechanical service company ran one dispatch system from 2011 to 2019, then moved to a new one. The migration carried over customers and open jobs. It left behind eight years of closed tickets, notes and estimates. The old system’s licence lapsed. Someone kept a backup.

This is where micro1’s wording helps most. It says companies with records “across multiple business systems are generally a stronger fit.” As our own reading, two systems are not a mess to apologise for. They are breadth. The old system holds the long run of history. The new one holds the recent years, often in more detail.

What matters is whether the old years can be opened. Our page on how many years of records count has the counting method, so we won’t repeat it. The point for this page is narrower. Don’t try to merge the two systems by hand into one clean set before you apply. A merge done in a hurry drops fields, breaks the links between estimates and jobs, and loses the notes. Describe each system and where the two overlap instead. Two honest exports, with the overlap named, beat one tidy file with holes in it.

Our advice: list both systems, the years each one covers, and whether you can still export from the old one. Note the overlap year. Then let the conversation decide what moves. Our page on whether the old system’s records still count shows how to keep them before the login lapses.

Paper, scans and handwriting

micro1 says nothing on its pages about paper or handwriting. It publishes no list of formats. Its only words on delivery are that you transfer the data “through our APIs or other approved methods.” As our own reading, what counts as approved is settled with each company.

So this section is our own advice. Boxes of carbon-copy tickets, signed estimates and handwritten inspection sheets still record how work was done. They are also slow to move. Scanning thousands of pages costs money and time. Typing them up changes them.

Before you scan anything, ask. Say how many boxes there are, which years they cover and what kind of form is in them. If micro1 sees a fit, its referral page says qualified companies are invited “to connect with a Data Partnership Manager to discuss their data, licensing terms, and next steps.” That is the place to find out whether paper is wanted and in what form. Don’t spend a month at the scanner on a guess.

Duplicates and copies

micro1 doesn’t mention duplicates either. This is our own advice too.

Most companies hold the same record more than once. The estimate PDF sits in the job folder, in the sent mail and in the shared drive. The old system and the new one both hold the open jobs from the switch year. Every revision of a quote may sit beside the final one.

Not every copy is a duplicate. Four versions of one estimate are a history of the negotiation. As our own reading, that is the kind of record micro1 describes as decision-making. Delete the older versions and you delete the reasoning.

True duplicates are different: the same file, unchanged, in two places. Don’t spend weeks hunting them. Duplicates left by a migration do weigh on quality, as our price page says. The fix is to tell micro1 where they are, not to delete them by hand. Its pages say datasets are “processed according to the agreed use case and handling requirements.” What gets cut is part of that agreement, not a chore you must finish first.

What is wanted, and what is not

All of this sits inside one rule that owners worry about most. The records are de-identified before use. micro1’s announcement says it “uses de-identified company data.” Its privacy list says “Personal and identifying information is removed or transformed before your data is shared with third parties.” Its protection answer says “Personally identifiable information (PII) is removed when relevant”. After sharing, it says, “certain datasets may undergo synthetic rewrites that preserve workflow structure while reducing direct ties to original records.”

So your customer names and your payments are never the point. As our own reading, a homeowner’s address and a card on file teach nothing about how a trade is run. What is wanted is the work itself. How a job was scoped. How an estimate became a job. How a change order was settled when the wall was opened. How a repair was diagnosed when the first fix didn’t hold.

That is why messy records can count. The names come out either way. The steps stay. A half-filled work order still holds the steps. Our page on how de-identification works shows, record by record, what comes out and what stays. Before prepared data is used, micro1 says companies are given “the opportunity to review representative samples.” Ask to see a sample drawn from your messiest records, not your cleanest.

What to tidy and what to leave alone

This list is our own advice, not micro1’s guidance.

Leave alone:

Blank fields. Don’t fill them in from memory.
Old revisions of estimates and quotes.
Abbreviations, shorthand and typos in notes.
The old system, as it is. Export it; don’t clean it.
Customer names. micro1 says you don’t need to anonymize first.

Do the work on:

Finding things. Which systems, which mailboxes, which drives, which boxes.
Counting things. Jobs per year, years per system, mailboxes per role.
Opening things. Can you still export from the old software? Does the backup restore?
Scoping things. Which jobs fall under client contracts that forbid sharing.
Describing things. micro1’s referral page says the company “completes data manifest” when it sets up its profile. A short, honest note on where each record lives and what shape it’s in belongs there.

Our page on what decides the price goes through micro1’s factors one at a time. As our own reading, none of them rewards a tidy-up done the week before you apply.

What this means for a company with 30 or more staff

micro1’s referral page lists “30+ employees” among its company requirements. Here is one company, as our own hypothetical: an HVAC and plumbing service firm with 55 people and fourteen years of records.

What it has. Work orders from two dispatch systems, split at a switch in 2019. Estimates that started in the software and finished in email. Photos on the job record from 2020 onward. Paper tickets for the first three years. A shared quotes inbox.

What looks messy. The old system’s cause field is blank on many tickets. The final price of half the larger jobs sits only in email. The paper years are in boxes.

What it does. Exports both systems as they are. Counts the mailboxes and years. Notes the paper boxes and their years without scanning them. Lists the client contracts that restrict sharing. Writes it all into the data manifest in plain words.

What it asks on the first call. Whether paper is wanted. Whether both systems should be sent whole. Which sample of prepared records it will be shown before use.

Our page on companies with 30 or more staff covers the lane in full. For a sense of scale first, our estimate gives a range in a few questions. It is our range, not an offer. Our page on what our estimate measures lists what it counts and what it can’t see, such as how complete a job file is.

What is never included

micro1 doesn’t publish a fixed exclusion list. It says it protects “sensitive information through data anonymization and secure handling,” and, as we read its pages, scope is agreed with each company.

As our own advice, keep these out unless micro1 agrees otherwise in writing: attorney-client privileged records, health information, payment-card and financial-account details, Social Security numbers, minors’ data, biometric data, and classified or export-controlled material. Leave out anything a client owns under its contract with you, and the customer list itself.

Messy records make this harder, not easier. A card number can hide in a free-text note. A tenant’s medical reason for an urgent call can sit in a work order. Our page on what is never included shows where those records tend to hide. Our How it’s safe page gives the short version of the protections.

Where we fit

We, TrustPast, explain how this works and give you a range with our estimate. You apply to micro1 yourself, through the button below. micro1 pays us a referral fee. You pay us nothing. We never hold your records, and we don’t tidy them for you. As our own advice, you don’t need to tidy them either.

TrustPast

Questions and answers

Does micro1 want only clean, complete records?

micro1 doesn’t say it wants clean records. Its announcement says “Real business operations involve incomplete information, competing priorities and exceptions that require judgment.” It does name “data quality” among the factors behind compensation, and on delivery it says “We review the data for completeness.”

Do we have to tidy or anonymize the records before we send them?

Not the names. Asked whether a company must anonymize its data before transferring it, micro1 answers: “No, micro1 has processes for de-identifying and transforming sensitive information.” As our own advice, don’t rewrite or fill in old records either. Find the missing systems and years instead.

Do records spread across several systems still count?

micro1 says “Companies with several years of records across multiple business systems are generally a stronger fit.” It also says “We evaluate the historical depth, breadth, quality, and relevance of the available data.” As we read it, being spread out is not a fault in itself.

What does micro1 say about handwritten files, scans or duplicates?

Nothing, on the pages we fetched. micro1 says “You securely transfer your data through our APIs or other approved methods.” It publishes no list of accepted formats. As our own advice, describe paper, scans and duplicate copies honestly in the data manifest and ask before you scan or merge anything.

Is there a minimum amount of records?

micro1 says: “There's no fixed minimum number of files, documents, or terabytes.” Final payout, its payout page adds, “is determined based on data sources, volume, quality, access terms, and due diligence.”

Does micro1 want our customer names and payments?

No, as we read micro1’s pages. It says “Personal and identifying information is removed or transformed before your data is shared with third parties.” What it looks for is “data that reflects how work is actually performed within an organization.”