The job that appears after go-live
Every CRM generates a steady trickle of work once the project team leaves. Users ask why a filter shows nothing. A new hire needs onboarding into the setup. Sales leadership wants a report changed. Someone requests a new field, and someone must have the standing to say no. Ignore this trickle and the account decays the way we describe in the data hygiene guide: duplicates, dead deals, fields nobody fills, reports nobody trusts.
The question is not whether this work exists. It is who does it: a named internal admin, an external partner on a retainer, or the combination. Each model has a real cost and a real failure mode.
The in-house admin model
The internal admin is usually a sales ops person or an operations-minded senior rep who picks up two to four hours per week of CRM ownership. The strengths are context and speed. They sit in the sales meetings, they know why stage three exists, and they fix small things the same day. Nobody external ever matches that latency.
The costs are subtler than the salary hours. Depth is one: an internal admin sees one Pipedrive account in their career, so novel problems, integration architecture and platform changes hit them cold. Continuity is another. When the admin leaves, the knowledge often leaves whole, undocumented. And mandate is the third. An admin without authority to refuse field requests becomes a ticket queue, and the field list grows back within quarters, exactly the pattern from common implementation mistakes.
The partner model
A partner on retainer brings pattern knowledge across dozens of accounts. They have seen your problem before, they know which platform updates matter, and their work survives your staff turnover. At €90 to €150 per hour, a modest retainer often costs less than the loaded salary hours it replaces, especially when the alternative admin is a senior rep whose selling time is worth more than the fee.
The failure mode is distance. A partner does not sit in your Monday meeting, so small frictions go unreported until they calcify. Response times are contractual, not immediate. And a bad partner builds quiet dependency: undocumented automations only they understand, priced accordingly. The tests in how to choose a partner screen for exactly that behaviour.
The hybrid most teams land on
After a year, most healthy accounts converge on the same shape. An internal owner handles the daily layer: user questions, small changes, hygiene routines, first-line judgement on requests. A partner handles the deep layer: integrations, automation architecture, quarterly reviews of the setup against how the business has drifted. The internal owner is the partner's counterpart, not their replacement, and each side keeps the other honest.
The economics work because the two layers price differently. Daily latency is cheap to staff internally and expensive to contract. Deep expertise is expensive to grow internally and cheap to rent. Buying each layer where it is cheap is the whole trick, and the contracting side of it is covered in project versus retainer.
How to decide for your team
Under ten seats with light integration needs: an internal power user with a partner on call for incidents is enough. Ten to thirty seats: the hybrid, with a small monthly retainer and a mandated internal owner. Above thirty seats, or with revenue-critical integrations: the internal role becomes real sales ops headcount, and the partner shifts to architecture and review work. If you are still pre-implementation, settle the build question first in DIY versus hiring a partner, then return here for the year-two staffing.