Directory · buyer guidance

Choosing an implementation consultant

For partner-implemented products — Acumatica, Business Central, Sage Intacct, NetSuite — the partner determines the outcome more than the product does. The variance between a good and a poor one exceeds the variance between the products they implement.

Reviewing a partner proposal?

Send it over. We will tell you which lines are soft and what to ask before signing.

1 / 3
We are not a partner for any vendorNo referral feesWeights from 41 stalled projects

Before you sign

Six questions, and what the answers mean.

Each of these maps to a primary cause of failure in the forty-one stalled implementations we were brought into. They are not general diligence — they are the specific things that went wrong.

Who specifically will be on our project?

By name, with their availability. The team that sold the project is frequently not the team that delivers it, and partner turnover was the primary cause in 8% of the stalled projects we examined.

What was your median variance to original quote last year?

A partner who tracks this and will say it is a different kind of firm from one who deflects. Our own sample puts the median at +27%, so an answer near zero deserves a follow-up question.

Give us two references whose projects overran.

Vendor-supplied references are selected. Asking for the difficult ones separates partners immediately, and the ones who provide them are usually the ones worth working with.

How will requirements be derived?

From interviews, or from our transactions? Unsettled requirements were the primary cause in 31% of stalled projects, and interview-only requirements are the most common reason they move.

What happens when the data is worse than expected?

It will be. The answer should be a defined process and a commercial position, not surprise. Data problems were the primary cause in 24% of stalled projects and were nearly always found during load.

Who on our side needs to be able to decide?

A partner who cannot answer this has not thought about the failure mode that accounts for 19% of stalled projects — a sponsor who attends everything and cannot decide across departments.

What the answers actually tell you

The content of the answers matters less than whether the partner has a considered position at all. A firm that has thought about its own variance, its own staffing risk, and its own failed projects is a firm that has processes around them.

Deflection is the signal. “Every project is different” in response to the variance question means either they do not measure it or the number is bad. Both are worth knowing before signing rather than in month four.

Ask for two references whose projects overran. The reluctance, or the willingness, tells you more than either reference call will.

Named people, with availability

The single most common gap between the sales process and delivery is staffing. The partner’s best consultant is in the room for the pitch and on three other projects by the time yours starts.

Ask for names, ask what else those people are committed to during your timeline, and put the named team in the contract with a clause covering substitution. Partners who resist that are telling you the pitch team was not the delivery team.

Certification is a weak signal

Vendor certifications measure product knowledge, which is necessary and not the thing that goes wrong. Nearly three quarters of the stalled projects we examined failed on requirements, data, or decision authority — none of which certification tests.

Deployment count in the last twelve months, for companies of your size and shape, is a considerably better signal. A partner with forty certifications and four recent deployments is a different risk from the reverse.

Contract terms worth negotiating

Most buyers spend their negotiating effort on rate. The terms that actually protect you are elsewhere, and partners have more discretion on them than on price.

  • Named team with a substitution clause. Replacements require your approval and equivalent seniority.
  • Discovery-failure allocation. Anything the partner should have found during scoping is theirs to absorb; new requirements are yours. Write the distinction down.
  • A data-condition contingency. Agree in advance what happens if the data is worse than the diagnostic found, rather than negotiating it under pressure in month three.
  • Phase gates with an exit. The right to stop after discovery, having paid for discovery, is worth more than a rate concession.
  • Knowledge transfer as a deliverable. Documented and accepted, not assumed. A partner you cannot leave is a dependency you bought without pricing.
Our position on this, disclosed

We are not an implementation partner for any vendor and take no referral fees from any partner or vendor. We also compete with these firms — we implement our own product and we rescue other people’s stalled projects. Read this page knowing that. The check on it is that everything here maps to published failure data rather than to a recommendation to use us.

Questions

Common follow-ups.

Do you recommend specific partners?
No, and we take no referral fees from any. We give you the questions rather than a list, because partner quality is regional and changes with staffing faster than any list we published would.
How much does implementation cost?
Our pricing study puts the median at 1.4× first-year licence, with more than 2× variance between quotes for comparable scope. The implementation line has far more room in it than the licence line.
Is certification worth weighting?
Weakly. It tests product knowledge, and three quarters of stalled projects failed on requirements, data, or authority. Recent deployment count for companies your size is a better signal.
What if our partner is already struggling?
Address it early and specifically. Of the 41 stalled projects we were brought into, the ones that recovered fastest were the ones where somebody raised it in month three rather than month eight.
Are you biased here?
Yes — we compete with these firms and we rescue their failed projects. Everything on this page maps to published failure data rather than to a reason to use us, which is the check you should apply.

The partner matters more than the product.

For partner-implemented systems, the variance between firms exceeds the variance between the products they implement.