Who this is for, and who it is not for
An audit is for the account that works well enough to keep using and badly enough that nobody is sure why. Reports disagree. Reps say the data is wrong. Somebody added a dozen fields last year and nobody remembers why. A former admin left, and the automations are folklore. You need to know whether to repair, to rebuild or to leave it alone, and you want that answer before you spend money.
It is not the right service when:
- You already know the structure is wrong. Go straight to rebuild versus repair and price the rebuild.
- The account is a few weeks old. Audit it at the thirty day mark, when real behavior has shown up in the data.
- You want the data cleaned. An audit finds and counts the problems. Cleaning is a separate, scoped step, and the routine is on the data hygiene page.
The checklist you can run without us
Run it in the order below, because each group depends on the one before. Mark every line green, amber or red. Do not fix anything while you score: an audit that changes the account as it reads it cannot be compared with last quarter's. Take an export first, as a backup.
Pipelines
Everything else inherits from the pipeline. The method behind these checks is on pipeline design and stages and probabilities.
| # | Check | How to test it | Healthy looks like |
|---|---|---|---|
| P1 | Each pipeline matches a distinct selling process | List the pipelines and name the process each one serves | No two pipelines serve the same process |
| P2 | Stage names describe what the buyer has done | Read each stage name aloud and ask what event moves a deal into it | Every stage is an event, not a mood |
| P3 | Every stage has a written exit criterion | Ask a new hire to say when a deal may leave a stage | The criterion fits in one sentence and is checkable |
| P4 | Stage probabilities match observed conversion | Compare each probability with won and lost history for the last year | Probabilities were set from data, not left at defaults |
| P5 | Stale deals show up as stale | Check that deal rotting is set per stage and the flagged deals are real | Rotting deals are few and each has an owner decision |
| P6 | Lost reasons are a short list and every lost deal has one | Filter lost deals with an empty lost reason | Zero empty, five to eight options, no free text |
Fields
Field problems are the quietest and the most expensive, because every report and automation reads them. See custom fields for the design rules behind this group.
| # | Check | How to test it | Healthy looks like |
|---|---|---|---|
| F1 | Each custom field names the report, filter or automation that uses it | Go through the field list and write the consumer next to each one | No field without a consumer |
| F2 | Fewer than a quarter of custom fields are effectively empty | Count fields filled on fewer than 10 percent of recent records | Dead fields are archived, not left to confuse |
| F3 | Fields used for grouping are option lists, not free text | Open every field a report groups by and look at its type | Single option for anything you group on |
| F4 | Each piece of data lives on the right object | Look for deal fields that describe the company, and person fields that describe the deal | Company facts on the organization, deal facts on the deal |
| F5 | The deal field set is small | Count the custom deal fields | Roughly eight to twelve, each one used |
| F6 | Required fields are few and appear where the answer is known | List mandatory fields and the stage at which each is required | Nothing is mandatory before the rep can know it |
Automations
Every automation is invisible machinery that keeps running after the process changes. These checks mirror the maintenance routine on the automations guide.
| # | Check | How to test it | Healthy looks like |
|---|---|---|---|
| A1 | Every automation has a descriptive name and an owner | Read the automation list and ask who changes each one | Named trigger then effect, one person owns the list |
| A2 | The execution history shows no recurring failures | Scan the last 30 days of history for repeated errors | No failure repeats, and none is ignored |
| A3 | No automation triggers another in a loop or at cross purposes | Match every action against every trigger on the same object | Chains are short, named and deliberate |
| A4 | Customer-facing sends are few and come from the owner | List automations that email a contact and the sender on each | Owner-signed, low volume, with a working opt-out |
| A5 | No automation points at a renamed or retired stage or field | Search the list for every stage and field changed in the last year | No orphaned references |
| A6 | Webhooks and tokens belong to live systems | Inventory webhooks and API tokens, then match each to a system and a user | Every token has an owner who is still active |
Data
These are counts, and counts are what make an audit credible. Save each as a filter so the next audit takes minutes.
| # | Check | How to test it | Healthy looks like |
|---|---|---|---|
| D1 | Duplicate people and organizations are under control | Run merge duplicates for people and organizations and count the pairs | Near zero, with an import rule that prevents new ones |
| D2 | No open deal has a close date in the past | Filter open deals with an expected close date before today | Zero |
| D3 | Every open deal has a next activity | Filter open deals with next activity empty | Zero |
| D4 | Silent deals are rare | Filter open deals with no activity in 45 days | Near zero, each with an owner decision |
| D5 | People belong to an organization | Filter people with organization empty | Near zero |
| D6 | No active records belong to a deactivated user | Filter deals, people and organizations by owners who have left | Zero |
| D7 | Source is filled on open deals | Filter open deals with deal source empty | Zero, and the list of sources is short |
Reporting
A report is only as good as its definitions. See dashboards and reporting and forecasting for how to fix what this group finds.
| # | Check | How to test it | Healthy looks like |
|---|---|---|---|
| R1 | Each of leadership's top questions maps to one saved report | Write the three questions down and find the report that answers each | One report per question, owned by a named person |
| R2 | Reports agree with each other | Compare won revenue for one month across two reports | The numbers match, or the difference is explained |
| R3 | Reports group only by fields that are kept filled | For each report, count blanks on its grouping fields | Blanks are a rounding error |
| R4 | The forecast rests on honest close dates and probabilities | Compare the weighted pipeline of two quarters ago with what actually closed | The gap is known and trending down |
| R5 | Dashboards exist per audience and somebody opens them weekly | Check for a team, manager and leadership view, and ask who opened each last week | Three dashboards, each with a regular reader |
Permissions
Who may do what, and who may see what, are separate decisions. The reasoning is on permissions and visibility.
| # | Check | How to test it | Healthy looks like |
|---|---|---|---|
| S1 | Admin rights sit with two people at most | List users with admin rights | Two, both named, both still employed |
| S2 | Every active user is a current employee or contractor | Match the user list against the staff list | Departed users are deactivated, and their records reassigned |
| S3 | Visibility is company wide unless an incident justifies a limit | Review visibility groups and ask what each restriction prevents | Each restriction names the incident it prevents |
| S4 | Export and delete rights are limited to those who need them | Review each permission set for export, bulk edit and delete | Few people can export the whole customer list |
| S5 | Email sync and shared inbox visibility are deliberate | Check who sees whose synced email and why | A written choice, not a default |
| S6 | Integrations run under a user who will stay | Find tokens and connected apps tied to one person's login | Tied to a service user or a long-term owner |
How to read your score
Count the reds per group, then read the pattern, not just the total. A cluster of reds in data and nothing else is a repair: duplicates, stale deals and dead fields, fixable in place in a few focused days. Reds in pipelines or in field placement are structural, and structure cannot be bulk edited into shape. Reds in permissions and automation ownership are governance, and they are usually quick to fix and slow to keep fixed.
| Pattern | Reading | Next step |
|---|---|---|
| Reds mostly in data, a few in automations | Sound structure, dirty surface | Repair, then install a monthly routine |
| Reds in pipelines or in where data lives | Structural fault | Rebuild the weakest structural element, repair the rest |
| Reds spread over four or more groups | The account teaches the wrong process daily | Rebuild, with a process session first |
| Mostly amber, a handful of red | Healthy account with drift | Fix the reds, schedule a quarterly review |
Pair the score with the five question test on rebuild versus repair. If you want a quick, free read before you start the full checklist, the readiness assessment takes ten questions and gives a score on screen.
What a paid audit delivers
You can do all of the above yourself, and the page exists so that you can. A paid audit adds speed and an outside view of what has become normal to you.
- The scored checklist, with evidence. Every check run against your account, with the count or the example that justifies the colour.
- A findings list in plain language. What is wrong, why it matters to a report or a process, and what it costs to leave.
- A prioritized fix list. Sorted into fix now, repair and rebuild, with the order in which to do them so nothing is done twice.
- A recommendation with a price band. Repair or rebuild, with the band the work falls in, so you can budget before you commit.
- A scope you can take elsewhere. Written so that any competent partner can quote against it. That is deliberate.
What we need from your team
- Admin access to Pipedrive, for reading. We change nothing.
- A list of every connected system, form and email tool, and who set each one up.
- The three questions leadership asks of the numbers every month, in leadership's own words.
- One person who knows why the account is the way it is, for a conversation.
- An hour with two reps, so we see how the account is used and not only how it is built.
What it costs
We do not publish a stand-alone audit price. It is a short, scoped piece of work, and you see the scope and the price in writing before you commit. What the audit leads to is priced in the open. A repair typically runs $1,650 to $6,600 over two to five days, and a rebuild prices like a fresh implementation, which is where the project bands apply.
| Team and starting point | One-off project | Typical duration |
|---|---|---|
| Team of 2 to 5, no migration, one pipeline | $3,300 to $5,500 | 2 to 3 weeks |
| Team of 5 to 20, migration from a sheet or a light CRM | $6,600 to $13,200 | 4 to 8 weeks |
| Team of 20 to 80, migration from HubSpot or Salesforce | $13,200 to $27,500 | 8 to 14 weeks |
| Above 80 users, which sits outside the bands we publish | Quoted separately | Set per project |
Read the rest of the cost picture on the implementation cost guide. For common errors that audits keep finding, see common implementation mistakes.
When this is not the right choice
Do not pay for an audit when the free checklist above is enough. For a small team on one pipeline, a self-run pass catches most of it, and the money is better spent on the fix. Do not pay for one either as a way to delay a decision you have already made. And do not treat an audit as a one-off event: it holds value only if somebody re-runs the checks each quarter and acts on the counts.
What clients say
5.0 from 11 reviews on the Pipedrive Marketplace, for Sales Surge, which powers PipeDriven. Top 8 Pipedrive partner worldwide, July 2026. Word for word from the listing, and describing our work in general:
Working with Sales Surge has brought clarity and structure to our enterprise sales motion. Their approach is practical and execution focused, helping us improve pipeline discipline, stakeholder engagement, and outreach consistency.
Very happy with the support from Bart from Sales Surge. Bart really understood our needs and helped us improve our Pipedrive setup and automations in a practical and efficient way. Communication was always easy and quick, and he was very helpful throughout the process. Would definitely recommend them!
Read all of them on the reviews page.
Next step
Run the checklist first. If the score leaves you with questions, or with a list longer than you can act on, bring it to the call and we will sort it with you.
Last reviewed: October 2026. Prices are in US dollars and rounded.