They Are Hiring for the Thing They Are About to Buy
Your rep has been calling Kestrel Fabricating for two years. Right voicemail, wrong week, every time.
Last Tuesday, Kestrel posted four jobs. A second shift press operator. A maintenance tech, and the posting names the brand and tonnage of the press. A quality engineer, and the requirements list asks for IATF 16949 experience. And a buyer.
Read those four together and you know more about Kestrel's next twelve months than your rep learned in two years of calling. They are adding capacity on a specific line. They are chasing automotive work, which means their approved vendor list is about to be rebuilt around documentation their current suppliers may not have. And the person who rebuilds it starts in about six weeks.
None of that is a secret. Kestrel published it. On their own website, in plain English, with a link on the homepage.
WHAT THE POSTING ACTUALLY TELLS YOU
You are treating it as a hiring announcement. HR news. Not your department.
It is actually the earliest public statement a company makes about what it is about to spend money on. It goes up before the capex request clears, before the RFQ, before anybody returns a cold call.
Why it beats intent data: intent data tells you an account is “in market.” A posting tells you the shift, the machine, the certification, and the city. One of those changes what your rep says. The other changes who they call.
Here is the part I did not expect when I started looking at this. The people who take job postings most seriously are not salespeople. They are investors.
S&P Dow Jones Indices publishes an index built on job openings scraped from company career sites, weighted by job count instead of market cap, updated weekly. Their stated argument is that postings lead payrolls and lead the general health of the economy, and they run eleven sector versions of it. Meanwhile a 2020 study in Management Science found that changes in a firm's online job postings predict the next year's growth in headcount, SG&A, sales revenue, and operating earnings, roughly three months before any of it shows up in financial statements. The effect was strongest when the hiring looked like expansion rather than replacement.
So the finance world has been paying real money for this feed for a decade. Your competitor's sales team, in all likelihood, has never opened it.
The Translation Table
The scan is not the hard part. Knowing what a posting means for what somebody buys is the hard part. This is the whole play:
- Second or third shift roles at an existing plant. Capacity is going up. Consumables, MRO, and packaging demand rise before the headcount is even fully hired, because they stage material first.
- A buyer, purchasing agent, or sourcing manager. One of the strongest signals for a distributor. It means the person responsible for your category is either new or about to be. In their first ninety days, changing suppliers can look like improvement. Once they settle in, keeping the vendor list becomes the safer choice.
- Maintenance tech, and the posting names the equipment. That asset exists, it is running, and it needs parts. Their website will never tell you the model number. The posting does, because they need somebody who has actually worked on one.
- Quality engineer with a named certification. AS9100, IATF 16949, ISO 13485. They are entering a market with different supplier requirements, which means the approved vendor list gets rebuilt around documentation. If you already have the paperwork and the incumbent does not, that is the opening.
- Warehouse or logistics roles at an address you have never shipped to. New location. You may already have the account and not the site.
- Postings that stop. An account that quits posting for the function that buys from you is contracting, not expanding. Most people running this play only look at the positive direction, which is how you end up prospecting into an account you should be defending.
This Week, Try This: The Hiring Signal Scan
Two versions. Start with whichever one matches how much friction you can tolerate this week.
The free version, thirty minutes. Take twenty target accounts. Open the careers page on each one, copy every current posting into a document, and run the prompt below. That is it. No tools, no signup. It is tedious and it works, and it will tell you within an afternoon whether the signal is real for your market before you spend anything.
The version that survives past week one. The manual approach dies the second week, because nobody has thirty minutes every Monday. If you want this to be a standing tool your reps actually get value from, you need the postings to come to you.
We use TheirStack. It aggregates roughly 230 million job postings across 13 million companies from job boards, ATS feeds, and company career pages, crawled daily, with history back to 2021. The reason it matters for this specific build: they publish an MCP server, which means you connect it to Claude once and then ask questions in plain language. No API code, no CSV shuffling. Your Sales Ops person wires it up in an afternoon and your reps ask it things.
Free tier is 1,000 company credits. Paid starts at $49 a month. (We have no relationship with them and get nothing if you sign up, in case that was a question.)
Connected, the workflow is just conversation. “Pull every posting from the last 30 days for these 40 domains. Flag anything that indicates second shift, new equipment, a new quality certification, or a purchasing hire.” Then you feed what comes back into this:
You are analyzing job postings to find buying signals for a
B2B distributor or manufacturer. You will get two inputs:
a list of target accounts with the product categories we
sell or would sell them, and a set of current job postings
from those accounts.
Your job is to translate postings into what each account is
likely about to buy, and to hand a rep something specific
enough to open a conversation with.
STEP 1 - CLEAN THE SET:
Drop any posting placed by a staffing agency or recruiter
rather than the company itself. Drop postings that appear
to be evergreen reposts, meaning the same requisition
appearing continuously for more than 90 days. Say how many
you dropped and why. A company that always has 12 openings
is not a company that is expanding.
STEP 2 - READ FOR OPERATIONAL FACTS:
For each posting, extract only what is stated, not what is
implied:
- Shift referenced (first, second, third, weekend)
- Any equipment, machine, brand, or line named anywhere in
the posting, including in the requirements section
- Any certification, standard, or regulatory regime named
(AS9100, IATF 16949, ISO 13485, FDA, ITAR, and so on)
- Any software or system named (ERP, WMS, MES, EDI)
- Location, and whether it differs from the company's
primary address
- Function and seniority, with special attention to
purchasing, sourcing, supply chain, and planning roles
- Date posted
STEP 3 - TRANSLATE TO BUYING IMPLICATIONS:
For each account, write what the postings taken together
suggest about what they are about to need. Tie each
implication to the specific product categories from the
account list. State a confidence level of high, medium, or
low, and say what would raise it.
Rank the implication types this way, highest value first:
1. A purchasing or sourcing hire, which means the vendor
list is in play
2. A new certification requirement, which means the
approved vendor list gets rebuilt
3. Named equipment we can supply parts or consumables for
4. Added shift or added location, which means volume
5. General growth with no specific tie to what we sell
STEP 4 - WRITE THE OPENER:
For each account you rate medium or high, write one
question a rep could open a call with. The question must
reference the operational reality, never the job posting.
"Are you running second shift on the press line yet, or
still ramping" is correct. "I saw you're hiring a press
operator" is wrong and you should never produce it.
STEP 5 - FLAG THE OTHER DIRECTION:
Separately list any account whose posting activity has
dropped compared to the prior period, or that is posting
only for backfill in functions unrelated to growth. Label
this list "defend, do not prospect."
RULES:
- Quote the exact line from the posting that supports every
claim you make. If you cannot quote it, do not claim it.
- Never infer financial health, funding, or revenue from
hiring volume alone.
- Do not write anything that references the job posting as
the source of your knowledge. That output goes to a rep
who will read it out loud.
- If an account's postings support no implication relevant
to what we sell, say so plainly and move on. A short
honest list beats a long padded one.
One rule that is not in the prompt, because it applies to the human. Never tell the prospect where you got it. “I saw your job posting” turns a good insight into proof that you have been watching them, and it tells them the knowledge came from a scraper instead of from understanding their business. Reference the implication, not the source.
What comes back is a ranked account list where each name carries a reason and a first sentence. Not “this account is showing intent.” Something closer to “they named a 400 ton Aida press in a maintenance req, we stock tooling for that line, ask whether they are running it on two shifts yet.”
First run is under an hour. Run it monthly, not weekly. Postings do not move fast enough to justify weekly and your reps will stop reading it.
An Honest Take
Signal-based prospecting mostly fails, and not because the signals are bad. It fails because the signal gets treated as a reason to call instead of a reason to call about something. I have watched companies buy expensive intent data, get back a list of accounts supposedly in market, and hand it to reps who then open with exactly the pitch they were going to use anyway. The list changed. The sentence did not. A signal that does not change the first sentence out of a rep's mouth is just a more expensive alphabetical list, and a year later everyone concludes that intent data does not work.
The limit, since it is real: this works unevenly. Big manufacturers post constantly and in detail. A 40 person job shop posts twice a year through a staffing agency and says nothing. If that is your ICP, run the free version on twenty accounts before you spend a dollar, and be willing to conclude it does not fit your market.
The Bottom Line
Kestrel published its plans and nobody at your company read them. That is the prospecting story, and it is the smaller half of this.
Here is the other half. Run the same scan on your customer list. Somewhere on it is an account that just posted for a buyer, which means the person who has been ordering from you for six years is leaving, and their replacement inherits a vendor list they did not choose and no particular reason to keep it. You are not the prospect in that scenario. You are the incumbent about to get reviewed by somebody with something to prove.
And new buyers do not evaluate suppliers by reading the contract. They evaluate by placing three orders and seeing what happens.
Their new buyer will judge you on the first three orders.
He inherits a supplier list he did not build, with ninety days to show he improved something. What he can actually measure in that window is which vendors are easy to order from and which ones generate work.
Y Meadows runs order entry end to end: reads the PO in whatever format it shows up, matches customers and SKUs, applies your pricing, freight, and credit rules, posts it clean into your ERP. Want to see it run on the messiest PO one of your accounts sends?