Hourly ranges, fixed-price benchmarks by project type, and the questions that tell you whether a quote has been scoped honestly.
Hourly, an independent Excel consultant charges $75 to $200. An agency charges $150 to $350. Below about $40 an hour you are buying a generalist, which is fine for data entry and reliably disappointing for anything containing business logic.
But hourly is the wrong unit for most of this work. What you actually want to know is what a dashboard costs, and the answer is a range you can plan around.
| Project | What it covers | Typical range | Turnaround |
|---|---|---|---|
| Formula debugging | Broken lookups, error auditing, slow recalculation | $150–$400 | 4–8h |
| Data cleaning | Deduplication, normalisation, type fixing on one dataset | $200–$600 | 8h |
| Reporting dashboard | Pivot dashboard with slicers over a defined dataset | $400–$1,200 | 24h |
| Power Query pipeline | Repeatable import, join and transform from several sources | $500–$1,500 | 24–48h |
| Financial model | Three-statement or driver-based forecast with scenarios | $500–$2,500 | 24–72h |
| VBA / Apps Script tool | Automation with error handling and documentation | $750–$3,000 | 24–72h |
| Training | Live workshop tailored to your own files | $100–$250/h | Scheduled |
Our own pricing sits at the lower end of these ranges — from $200 for repair work, from $500 for a dashboard or model, from $750 for automation — because we quote fixed and deliver in 24 hours rather than staffing a project team.
This is the single biggest variable, and it is invisible in the brief. A dashboard over a clean table with consistent keys is a day's work. The same dashboard over an export where the customer name is spelled four ways, dates arrive as text in two formats and 3% of rows are subtotals masquerading as data is three days' work — most of it before anything visual gets built.
If you want a cheaper quote, the highest-leverage thing you can do is send a genuine sample of the real data rather than a tidied version of it.
"A sales dashboard" and "these six KPIs, filtered by region and month, in this layout" are different jobs. The second can be quoted precisely and built once. The first will be built twice, and someone pays for that.
A one-off analysis is cheap. Something that refreshes every month against a source you do not control needs validation, failure handling and documentation, and that is typically 30–50% on top of the build. It is also the part that gets cut from low quotes.
Building something only its author can modify is faster. If you need your own team to maintain it, the naming, structure and documentation have to be deliberate, and that is real time.
Hourly billing places the risk of a hard problem on the buyer. If a lookup fails because half the keys have trailing whitespace, that discovery takes an hour, and on an hourly contract you pay for it. On a fixed quote, the consultant absorbs it — which is why a fixed quote forces them to open your file before committing to a number.
The usual objection is scope creep, and it is a real risk in both directions. The answer is not hourly billing; it is a quote that says what is included and what a change costs. Ask for that explicitly.
Three cases where consulting is the wrong purchase:
✓ Fixed price before any work starts · ✓ Money-back guarantee
Common questions about consultant rates
Independent consultants typically charge $75–$200 per hour depending on seniority and region; agencies charge $150–$350. Rates below $40 usually mean offshore generalists, which works for data entry and rarely works for anything with business logic in it.
Fixed price, in almost every case. It moves the risk of an unexpectedly hard problem onto the person best placed to judge it, and it forces the consultant to read your file before committing. Hourly makes sense only for genuinely open-ended work such as ongoing support.
Usually because the scope is not the same. One quote covers the dashboard; another covers the dashboard, the data pipeline that feeds it and documentation. Ask each quote to state explicitly what happens when your source data changes shape — the answers will differ more than the prices.
For a defined dataset and an agreed layout, 8 to 24 hours of work. Longer estimates usually include cleaning the source data, which is real work but should be quoted as its own line rather than hidden inside the dashboard price.
Related guides and services