Drafted for review. Last source review: 11 October 2026. No affiliate links are used in this article.
A free trial is easy to waste. You click through a polished dashboard, create a sample quote, admire the calendar and decide that the software feels simple. Then a real customer changes the job, a subcontractor needs access, a phone loses signal, a certificate needs an attachment, or the accounting export does not contain the detail you expected.
The better approach is to treat a free trial as a controlled field test. Give the software a small, realistic workload, record what happened and decide against written pass or fail criteria. This guide is for UK trade businesses testing job-management, quoting, scheduling, invoicing, forms and field-service software. It is independent editorial guidance, not a recommendation of any provider. Free-plan terms, trial length, prices, limits and features can change, so check each provider's current UK page and contract before signing up.
The real aha is that the best free trial is not a product tour. It is a rehearsal of the moment when the business is busy, information is incomplete and somebody needs to find the truth quickly. Test that moment and the choice becomes less about which screen looks nicest.
The trial began with a quote that was never meant to be real
Imagine a heating business comparing two apps. The owner creates a fictional customer, adds a boiler replacement and sends a test quote to an internal email address. Both systems look capable. The owner chooses one and imports the customer list.
A week later, a real job exposes the missing questions. The customer accepts one option but not another. A subcontractor needs only the site details. The engineer adds photographs from a phone. The quote becomes a job, then a partial invoice, then a credit note. The owner needs the material cost and the latest customer message without exposing every customer record to every user.
The trial did not fail because the dashboard was poor. It failed because the test stopped before the business's information changed shape. A quote becoming a job, a job becoming an invoice and an invoice being corrected is where software either removes double entry or creates a new form of it.
Use the checklist below to test that movement. The aim is not to fill every feature with sample data. It is to prove the few workflows that carry money, customer promises, field evidence and operational risk.
Decide what the trial can and cannot answer
A software trial can tell you whether the product fits your workflow, devices, staff habits and budget assumptions. It can show how data is entered, moved, corrected, exported and retrieved. It can reveal which plan contains a feature and which usage limit affects the price.
A trial cannot prove that your VAT treatment is correct, that an electrical certificate meets the requirements for a particular job, that a customer will pay faster, or that the provider's security is appropriate without reviewing the contract and configuration. Software may support those activities, but the business remains responsible for its decisions and records.
Write the scope in one paragraph before creating the account:
We are testing this system for a small UK trade business with two office users, three field users, domestic and commercial jobs, quotes, variations, photographs, invoices, purchase costs and accounting export. We will use fictional customer data and will not invite live customers during the trial.
Change the paragraph if it does not describe your business. A sole trader testing a one-user free plan is not testing a five-person operation. A maintenance company testing one installation quote is not testing recurring work. A business that issues certificates needs to test the exact document process, not just job cards.
This boundary protects the trial from becoming a vague feeling. It also prevents a sales demo from answering a different question than the one the business actually has.
Write pass or fail criteria before you open the account
List the five outcomes the system must support. Keep them observable. `Easy to use` is too vague. `A field engineer can create a job from an accepted quote, add three photographs offline, record a variation and sync the record without duplicate entries` is testable.
Useful criteria include:
- a quote can become a job without losing customer, address, scope or pricing;
- a variation can be approved and shown separately from the original scope;
- staff see only the jobs and data appropriate to their role;
- field records work on the actual phones and tablets used by the team;
- photos, forms, certificates and notes remain attached to the correct job;
- deposits, progress invoices, credits and final invoices can be represented;
- costs, time and materials can be reported without manual reconstruction;
- accounting export retains the detail needed for reconciliation;
- records can be exported and read if the subscription ends;
- the total price remains acceptable after users, messages, payments, forms and add-ons.
Mark each criterion as `must have`, `useful` or `not relevant`. A trial should be allowed to fail. If every result is described as `probably fine`, you have collected impressions rather than evidence.
Build a five-job pack that behaves like your week
Create five fictional jobs using realistic but non-identifying data. Do not upload live customer names, addresses, phone numbers, access codes or photographs unless the provider's terms and your data protection assessment allow it. The ICO says organisations must keep personal data secure using appropriate technical and organisational measures, and that the right level depends on the risk.[17][18]
Your pack might contain:
- a small reactive repair with one engineer and a same-day invoice;
- a quoted domestic installation with options and a customer variation;
- a commercial job with several visits, purchase costs and a subcontractor;
- a recurring maintenance visit with a service history and reminder;
- a certificate or form-driven job with photographs, measurements and a customer handover document.
Use the same pack for each candidate. Record the time spent, the number of manual entries, the missing fields and the exported files. Do not compare a fully configured candidate against an untouched candidate. Give each product the same fair preparation allowance.
The pack should include at least one awkward event: a customer changes the scope, a supplier cost arrives late, an engineer cannot get a signal, a visit is cancelled, a payment is part-paid or a document needs correcting. The awkward event is often more informative than the happy path.
Start with the enquiry and follow the customer promise
Begin outside the main dashboard. Create an enquiry or customer request as your business receives it. Add the original wording, preferred dates, access details and source of the enquiry. Convert it into a quote and check whether the record retains the context or forces a fresh entry.
Create two quote options if your business sells choices. Test whether labour, materials, mark-ups, tax treatment, discounts and exclusions are clear. Send the fictional quote to an internal mailbox and open it on a phone. Check the document's legal and commercial details, including business identity, customer details, description, terms and expiry information.
Accept one option and reject the other. See what happens to the rejected option. Then create a variation after the work is scheduled. The software should make it possible to distinguish the original agreement from the later change. If the audit trail simply overwrites the original price, record that as a risk.
Ask whether customer acceptance is a typed response, a click, a signature, a payment or only a status change. Those events may have different evidential value. The software can help organise the record, but it does not decide the legal effect of a contract or signature.
Run the field test on the devices people actually carry
A product can feel excellent on a large office monitor and awkward on a phone with a cracked screen, a slow connection or gloves in the back of a van. Test on the operating systems and device sizes that the business will actually use. Use the camera, keyboard, notification settings and browser permissions as a field worker would.
Create a job from the accepted quote. Add notes, a photo, a document, a customer signature or acknowledgement and a time entry. Try correcting a spelling error without deleting the original context. Turn off the signal or use the provider's documented offline mode. Save the record, close the app and reconnect. Check for duplicate jobs, missing photos, conflicting edits and misleading timestamps.
Test the small details that determine whether people will use the system:
- Can an engineer see the address without opening unrelated customer data?
- Can they call the customer without copying the number elsewhere?
- Can they add a material that was not in the quote?
- Can they record a reason for delay or a return visit?
- Can they find yesterday's notes from the same site?
- Can they send the correct document without accidentally sending an internal note?
- Can they work when the customer wants a change before the office is available?
Do not confuse fewer taps with better control. A shortcut that removes the record of who approved a variation may be convenient but unsuitable for a business that needs a clear audit trail.
Test the money trail from price to payment
Use the five-job pack to test the commercial flow. Record labour, materials, subcontractor costs, deposits, part payments, retention, discount and a credit or correction where the business uses them. Check how the software distinguishes an estimate from a quote, a quote from an invoice and an invoice from a payment request.
The screen should make it clear what has been charged, what is outstanding and what is still only an internal cost. If a product uses a job value that changes whenever someone edits an estimate, see whether the original approval remains available.
For VAT-registered businesses, test the accounting treatment with your accountant's approved example data. A free trial is not the place to let a default rate decide a real tax question. Check whether the product supports the rates, reverse-charge scenarios, credit notes and dates that your business actually uses. HMRC record-keeping guidance expects complete business records and, where relevant, digital records in functional compatible software. An integration badge does not prove that your workflow preserves the required detail.
Export one sale, one purchase, one correction and one payment. Reconcile the export to the app screen and to a simple control sheet. Record any field that disappears or changes meaning. If a job-management system sends only a summary to accounting software, document where the detailed source record remains and how the two systems are linked.
Treat forms and certificates as evidence workflows
A form button is not the same as a compliant document. If your trade uses inspection records, service reports, risk forms, certificates or customer handover documents, test the entire workflow. Start with a blank record, assign it to the right job, complete it on the field device, add supporting evidence, correct a mistake, obtain the intended approval and issue the final version.
Open the exported document without the app. Look for the business identity, customer and site, work description, responsible person, date, relevant measurements, limitations, attachments, document number and correction history. Check whether the customer receives only the intended copy. Check whether the business retains the source data and the final issued document.
For electrical work, gas work, water systems or other regulated activity, do not treat a provider's statement that it supports certificates as a decision that your form is suitable. Compare the output with the current government, regulator, scheme or professional requirements relevant to the work. The existing software guide can make record handling easier, but the business still needs the right competence, inspection, testing and notification route.
If the app offers calculations, ask what inputs are required, what assumptions are built in, how the result can be checked and how a changed input is recorded. A calculation result without its inputs is a weak record.
Check people, permissions and personal data before inviting the team
The ICO distinguishes between controllers and processors and says the relationship should be clear. A cloud service processing customer data on behalf of a trade business is commonly considered as part of that controller and processor assessment. Review the provider's data-processing terms, security information, sub-processors, deletion and return arrangements before loading customer data.[17][19]
During the trial, use fictional accounts for the office manager, field worker, subcontractor and owner. Attempt to view, edit, export and delete information with each role. Test whether a subcontractor can see another customer's address, internal margin, supplier cost or unrelated photographs. Test what happens when a user leaves.
Ask the provider:
- where the data is stored and whether transfers outside the UK are involved;
- how access is authenticated and whether multi-factor authentication is available;
- how devices and sessions can be revoked;
- how backups, deletion and restoration work;
- how long data is retained after cancellation;
- how personal data is returned or deleted at contract end;
- which sub-processors handle messages, payments, storage, analytics or AI features;
- how a security incident is communicated.
The ICO says appropriate security includes confidentiality, integrity and availability, and that organisations should consider restoration and testing of measures.[17] The trial cannot certify the provider's security. It can reveal whether the provider gives enough information for an informed decision and whether the product's controls fit the business's risk.
Price the real account, not the attractive headline
Trial pages often present a starting plan. The starting figure is not necessarily the cost of the workflow you tested. Count every input that affected the pack:
- number of full and limited users;
- office and field access;
- forms, certificates and document storage;
- messages, calls and payment processing;
- accounting integrations;
- additional locations, assets or recurring work;
- onboarding, training and support;
- data import and migration;
- usage caps, job credits and overage charges;
- annual commitment, monthly pricing and renewal terms.
Provider pages show how offer structures can differ. ServiceM8's UK page describes a free plan with one user, a monthly job allowance, included features and a free trial with no card required.[13] Tradify's UK pages currently advertise a 14-day trial, show separate plans and list plan-specific features, add-ons and per-user pricing.[14] Fergus's UK pricing page currently advertises a 14-day free trial, no card required and different plan feature sets.[15] These are provider statements at the time checked, not guarantees for every business or a recommendation. Confirm the current offer, UK currency, tax treatment, limits and contract before entering billing details.
Jobber's UK pricing page invites a free trial but presents its product and plan information as part of a broader sales journey.[16] That is a reminder to capture the exact commercial terms shown at sign-up, not only the headline on a landing page.
Save a PDF or screenshot of the pricing and trial terms you relied on, with the date and URL. Do not assume the offer you saw in a search result is the offer attached to your account.
Put cancellation and exit in the test calendar
A trial is incomplete until the business can leave cleanly. Put a cancellation reminder in the calendar when the account opens. Find the cancellation control before adding any data. Record whether cancellation is self-service, whether the trial turns into a paid subscription, whether billing details are required, and what happens to the workspace afterward.
Create a small exit pack before the trial expires:
- customer and supplier records in a portable format;
- quotes, accepted variations and invoices;
- job notes, photos, forms and certificates;
- payment and cost information;
- user and permission list;
- integration settings and mapping notes;
- provider support replies and commercial terms.
Open the exports on a separate device. Check filenames, dates, attachments and readability. A CSV with no photographs and a PDF with no underlying measurements may be insufficient for your operational or compliance needs. Ask how to export data after cancellation if the trial does not expose the function.
Do not wait until the final hour to discover that a free account cannot export the records you created. The ability to leave is part of the buying decision, not a post-purchase nuisance.
Let a field worker try to break the workflow
The owner should not be the only tester. Give a field worker one fictional job with minimal explanation and observe what they do. Do not rescue them immediately. Note where they hesitate, which labels they misunderstand, where they create duplicate information and what they write in a free-text box because a required field is missing.
Then give an office user the same record and ask them to prepare an invoice, find the latest customer communication and identify what remains outstanding. Compare their understanding with the field worker's intention. A good workflow reduces translation between office and site. A poor workflow moves confusion from one person to another.
Ask the tester three questions:
- What would you avoid entering when the day is busy?
- What would you record somewhere else?
- What would you need to ask the office to do?
Those answers are evidence about adoption risk. A feature can exist and still be unusable if it requires a sequence that does not match the working day.
Make the decision with evidence, not relief
At the end of the trial, score each must-have criterion as pass, partial or fail. Add the evidence location, such as `exported job 03 PDF`, `pricing capture dated 11 October` or `provider reply in trial mailbox`. A partial result needs a named workaround, owner and cost. If the workaround depends on one person's memory, treat it as a risk rather than a solution.
Use a simple weighted view:
| Area | Question | Result | | --- | --- | --- | | Field use | Can the team complete a real job on its devices? | Pass, partial or fail | | Customer record | Does the history remain connected from enquiry to payment? | Pass, partial or fail | | Commercial control | Can the business handle changes, costs and corrections? | Pass, partial or fail | | Evidence | Can forms, photographs and issued documents be retrieved? | Pass, partial or fail | | Data protection | Do roles, contracts and security information fit the risk? | Pass, partial or fail | | Exit | Can the business export and cancel without losing control? | Pass, partial or fail | | Cost | Is the tested workflow affordable at expected volume? | Pass, partial or fail |
Do not let a low monthly price compensate for a failed exit test or missing field record. Do not let a beautiful interface compensate for an accounting export that cannot be reconciled. Do not let a free trial create a commitment before the business has reviewed the terms.
The best trial produces a smaller shortlist
A disciplined free trial may eliminate a product quickly. That is useful. The goal is not to make every provider look good. The goal is to find out which system can carry your business's actual information without hidden manual work, unclear ownership or unacceptable exit risk.
Use fictional data, five realistic jobs, real devices, an awkward event, a money trail, a document workflow, role testing, an export and a cancellation check. Read the current provider terms and the ICO guidance where customer data is involved. For tax, certificates and regulated work, test the record path but obtain professional confirmation for the underlying obligation.
The final question is practical: if tomorrow becomes messy, will the system help the business prove what happened? If the answer is evidenced by the trial pack, you have learned something valuable. If the answer depends on a sales promise or a screen you never exported, keep testing or walk away.
Affiliate status: no affiliate links are used on this page. Provider links are direct sources only, pending programme approval.