Somewhere between a request and a decision, your process leaves the system and becomes a shared workbook, a chain of forwarded e-mails and someone chasing. It works, mostly, until an auditor asks who approved a particular thing and on what basis — and the answer has to be reconstructed from a mailbox.
Replace our spreadsheet-based approval process.
A spreadsheet is an excellent modelling tool and a poor system of record. The moment more than one person needs to act on it in sequence, it stops being a document and starts being a workflow — but it has none of the properties a workflow needs. There is no single current version, no concept of whose turn it is, no clock, and no history.
The e-mail half is worse, because it looks like it is working. Approvals genuinely happen and decisions genuinely get made, so nobody escalates it. What is actually missing only becomes visible under scrutiny: whether the right person approved it, whether they saw the version they think they saw, whether anyone approved something outside their authority, and how long any of it took.
The other cost is invisible until it is measured. Nobody can improve a process they cannot see, so the queue that adds nine days is indistinguishable from the one that adds two, and effort goes wherever the loudest complaint is.
A workflow with as many stages as the process really has, where each stage assigns particular fields to particular people and controls which earlier fields they can see or change. The request is one record moving through stages rather than a document moving between people.
Approval itself is an electronic authorisation step: the approver receives an e-mail and approves or declines by link, without needing to log in and find anything. That single detail is usually what decides whether the system is adopted, because approvers are the least captive audience in any organisation.
Routing is conditional, so value thresholds, departments and exception cases send a request down the right path automatically rather than depending on the requester knowing the rules. Scheduled tasks chase what has gone quiet, and every step is timestamped and attributable, which turns the audit question from an exercise into a report.
The columns become a datastore with real field types and validation, so a date is a date and a cost centre is a value that exists. If the spreadsheet is the starting point, easy-create mode gets it in as it stands and you tidy from there.
A conditional form, so the fields shown depend on what is being requested rather than presenting everyone with every field. Dynamic add-fields handle the cases with a variable number of lines, and save-and-resume covers the requests that need someone to go and find a number.
Unlimited stages, each assigning its own fields to its own people, and each able to control which earlier fields stay visible or editable. This is where the rules that currently live in people's heads get written down — usually the most valuable conversation in the whole project.
Each stage can e-mail an authorisation request; the recipient approves or declines by link and the record moves on. Declines return the request to the requester with the reason attached rather than ending it.
Scheduled tasks nudge what has been sitting too long and escalate what has been sitting far too long. Because every transition is timestamped, the reports that follow show where the time actually goes rather than where everyone assumes it goes.
Nothing here is written specially for this use case — it is the same platform every ActiveManage application is built from. The full feature list is on the platform page.
| Feature | What it does | Why it matters here |
|---|---|---|
| Workflows | Unlimited stages, each assigning different fields to different people, with control over which earlier fields remain viewable or editable. | A request is one record with a current stage and an owner, instead of a document whose latest version is a matter of opinion. |
| Electronic authorisation | A stage e-mails an authorisation request; the recipient approves or declines by link, letting the submission proceed. | Approvers are usually senior and always busy. Anything that requires them to log in and go looking is where adoption dies. |
| Conditional form builder | A drag-and-drop builder for forms whose field sets change based on the options chosen. | Routing rules stop depending on the requester knowing them. Choosing the category picks the path. |
| Auditing | Every add, edit and delete recorded with who, when and from which IP, with rollback to any previous version. | This is the entire answer to "prove who approved this and what they were looking at when they did". |
| Scheduled tasks | Automated tasks at any interval. | Chasing becomes something the system does rather than a job on somebody's list. |
| Reports & graphs | A report builder that joins related tables, with filters, date ranges and chart output. | Cycle time per stage, in a chart. Usually the first time anyone has seen where the delay actually is. |
One process, end to end, with real requests going through it. Approval workflows are best proven narrow — a single well-chosen process converts sceptics faster than a broad rollout.
The remaining processes, which are usually much faster because the patterns, permissions and e-mail templates already exist.
Adding a stage, changing a threshold or adding an approver is configuration. This is the category of system clients most often end up running themselves.
No. A stage can e-mail an authorisation request and the approver approves or declines by link. Where the decision needs more context, the link can take them into the record — but the common case does not require an account or a password.
Yes. Conditional routing on any field, so thresholds, departments, categories and exception cases each take their own path without the requester needing to know the rules.
The request returns to the requester with the reason attached, and the decline is part of the permanent history rather than being overwritten when it is resubmitted.
Yes — that is the main reason people build these. Every transition records who, when and from which IP address, the full change history is retained, and you can roll back to see the record exactly as it stood at any point.
A single well-defined process is usually weeks rather than months. The variable is almost never the software; it is how long it takes your organisation to agree what the rules actually are.
Tell us what you are trying to fix and we will tell you whether this is the right shape for it — including when it is not.