The comparison usually gets made on salary alone, which is why it usually gets made wrong. An SDR is not a salary line; it is a salary plus tooling plus ramp plus management attention. Once all four are on the page, the decision changes for a lot of small teams.
In-house SDR: the real annual cost
- Base salary plus commission: varies widely by market
- Employer taxes and benefits: typically 15 to 30 percent on top
- Tooling per seat: CRM, sequencer, data provider, dialer, LinkedIn licence
- Recruiting cost: agency fee or your own hours screening
- Ramp: 60 to 90 days before consistent qualified meetings
- Management: a manager or founder spending real hours weekly on coaching and pipeline review
Outsourced SDR support: the real cost
- Monthly retainer, no employer taxes or benefits
- Tooling often shared or supplied, not per-seat new spend
- No recruiting cycle, so start is weeks not months
- Shorter ramp, because the person has done the motion before
- Lower management load if the engagement has one accountable owner
When in-house genuinely wins
- You already have a validated, repeatable motion worth scaling with three or more reps
- Your product needs deep technical discovery on the first call
- You have a sales manager with capacity to coach weekly
- The role is a deliberate career pipeline into AE
When outsourced wins
- You are still testing a segment and do not want a 12-month commitment
- Prospecting stops every time the founder gets busy closing
- You need pipeline before you can justify headcount
- Your AEs are strong closers who are being wasted on list building
The hybrid most small teams should actually run
Outsource the top of funnel: research, list building, sequencing, first-line objection handling, and qualification. Keep discovery and closing in-house. This gives your closers a full calendar without handing the customer relationship to someone outside your team.
How to judge either option in 60 days
Agree the qualification bar in writing before day one, and measure four things: qualified meetings booked, show-up rate, meeting-to-opportunity conversion, and reply quality by sequence. If any of those are unmeasured after 60 days, the problem is the setup, not the hiring model.