I pulled 91 remote job posts to price an outsourced SDR

I pulled nine remote job feeds on 19 September 2026 and counted what was in them. Not to make a point about hiring — to answer a question buyers keep asking in a way neither side can win: is an outsourced SDR cheaper than hiring one?

The answer is not a price. It is a denominator.

What the job feeds actually showed

The pull was WeWorkRemotely's general feed plus its eight category feeds: 91 unique posts carried the company's own domain in the post body. 65 of those domains answered with HTTP 200. Of the 65 live ones:

• 19 were sales, growth or marketing roles — demand generation, inside sales, heads of marketing and sales, affiliate and lifecycle marketing;
• 3 of those were revenue leadership titles, i.e. a company that has already decided to staff the function and is only choosing how;
• across the entire pull, exactly two posts printed money: $40,000/year (a support role) and $170/hour (a contractor).

That last line is the market's real condition. Almost nobody publishes a price, so buyers compare shapes instead: who writes the list, who owns the copy, what gets reported, what happens when it stops.

The denominator nobody quotes

A strict two-variant test at a 3% base rate with a +20% relative lift needs 13,914 contacts per arm — 27,828 in total — at 80% power and alpha 0.05. At 500 contacts a day that is 56 days. At 1,000 a day, 28 days.

So the only comparable number between two quotes is:

cost per finished reading = (fee + data + tooling + mailboxes) / verdicts per quarter

Two vendors and one verdict means the second invoice bought nothing. Adding a person does not change the number of readings either, until the volume itself changes — below roughly 2,000 contacts a day, the constraint is the funnel, not the staffing.

Six rows, of which most services sell three

The failure mode is a service that sells the first three rows, reports on them, and leaves the last two empty. You end up owning a dashboard of activity and no verdict.

Ten questions that separate two SDR quotes

1. Who is the named person on my account, and how many accounts do they hold?
2. What is a qualified conversation, in words I can point at? ("A reply" is not an answer.)
3. Which number does the channel own, and what volume does it need before that number can be read?
4. What is the stop rule — at what cost per qualified conversation do we stop?
5. Who owns the domains, mailboxes and sending reputation if we part ways?
6. Who owns the list, and can I export it with verification results and dates?
7. Are data, tooling and mailbox costs inside the fee or outside it?
8. What is the reporting unit — sends, replies, meetings, or finished readings with volume attached?
9. What does the first week produce that I keep even if we stop on day seven?
10. Which of my definitions did you have to invent because I never wrote them?

Question 10 is the useful one. The honest answer is a list of the things the engagement should have started with.

What the first week should hand over

Four artifacts, keepable even if the engagement ends immediately: the qualified-account definition, the number the channel owns, the volume that number needs before it can be read, and the stop rule. Everything after that is execution, and execution is cheap to replace.

I run this on my own domain, which is why the numbers above are countable rather than claimed. If you want the same page of definitions written for your funnel before you take another quote, that is the fixed-scope entry point: marketing engineering engagements, $900 Sprint / $1,900 Engine / $2,900 Full Build.

The canonical version of this piece, with the full table of the feed pull and the FAQ, is on my site: Outsourced SDR: what you are actually buying.