Hunch

Classify job titles in Google Sheets

Updated · Edoardo Panichi, maker of Hunch

=HUNCHSCORE(text, question, levels) rates a job title along an ordered scale, individual contributor to c-level, and returns the weighted position. =HUNCHPICK(text, options) reads the same title and returns a department, so a column of messy titles turns into two columns you can sort and filter a contact list on.

Why "VP" is not a rule

A rule that reads for the word director catches "Director of Sales" and misses "Head of Sales", "Sales Lead", and "Regional Sales Manager", four titles for roughly the same seniority. Seniority is a judgment about the whole title, not one word inside it, which is why a score works where a keyword list runs out of synonyms.

Seniority

=HUNCHSCORE(A2:A1000, "How senior is this person based on their job title?", "individual contributor|manager|director|vp|c-level", "label")

In Excel: =HUNCH.SCORE(A2:A1000, "How senior is this person based on their job title?", "individual contributor|manager|director|vp|c-level", "label")

The fourth argument, "label", returns the level as a word. Drop it and the formula returns the weighted position from 0 (individual contributor) to 4 (c-level), which is what you want for sorting a roster or averaging a team. Full syntax is on the HUNCHSCORE reference.

Department

=HUNCHPICK(A2:A1000, "engineering: builds or maintains the product|sales|marketing|customer support: support and success roles, not sales|finance and accounting|hr and people|operations|product: product management and design, not engineering|other")

In Excel: =HUNCH.PICK(A2:A1000, "engineering: builds or maintains the product|sales|marketing|customer support: support and success roles, not sales|finance and accounting|hr and people|operations|product: product management and design, not engineering|other")

Two of the options carry a short description after the colon, because customer support and sales both show up in titles that mention accounts, and product and engineering both show up in titles that mention software. Option limits and the confidence mode are on the HUNCHPICK reference.

Real outputs

Nine titles pulled from a CRM contact export, exactly as they were typed by whoever filled in the field.

RowABC
1Job titleSeniorityDepartment
2VP of Customer Experience & Successvpcustomer support
3Sr. Software Engineer IIindividual contributorengineering
4Head of Growth Marketingdirectormarketing
5Chief of Staff to the CEOvpother
6Accounts Payable Specialistindividual contributorfinance and accounting
7Regional Sales Manager, EMEAmanagersales
8Co-Founder & CEOc-levelother
9People Ops Generalistindividual contributorhr and people
10Customer Success Associateindividual contributorcustomer support

The department pick is doing real work, not keyword matching: "People Ops Generalist" lands on hr and people rather than operations, even though the word ops sits right there in the title, because the role description points at people, not the word. Two titles carry no clear department at all, Chief of Staff to the CEO and Co-Founder & CEO, and both land in other rather than being forced into a function they do not belong to, which is the option list working as designed.

Titles with no clean signal

Real exports carry titles that were never meant to be read by anyone outside the company. These six are the kind that break a keyword rule.

RowABC
1Job titleSeniorityDepartment
2Ninja of Awesome, Customer Happiness Teamindividual contributorcustomer support
3Founderc-levelother
4Intern, Marketingindividual contributormarketing
5Director-elect, pending board approvaldirectorhr and people
6Freelance Graphic Designer (contract)individual contributorproduct
7Business Developmentindividual contributorsales

A bare "Founder" scores c-level with nothing else in the cell, the same result as Co-Founder & CEO above, confirming the FAQ answer below: seniority reads the word, not the headcount, so a real one-person shop needs the size added to the text if that distinction matters to you. "Freelance Graphic Designer" lands on product rather than marketing, because the department list defines product as management and design; that is the option list doing exactly what it was written to do, though it is worth rereading if your team expects designers to land in marketing instead. "Director-elect, pending board approval" gives the department pick nothing to work with, and the hr and people it returned should be read as a low-confidence guess, not a signal, since the title names no function at all.

Sort and group the roster

=SORT(FILTER(A2:C1000, C2:C1000="engineering"), 2, FALSE)
=COUNTIFS(C2:C1000, "sales", B2:B1000, ">=3")

The second formula counts VP and c-level titles in the sales department, the rows worth mapping by hand before an outbound push. Swap sales for any department and ">=3" for whatever seniority floor matters for the list.

=AVERAGEIFS(B2:B1000, C2:C1000, "engineering")

Average seniority by department once both columns exist. A department that averages under 1 is mostly individual contributors, useful context before deciding who gets an executive outreach sequence and who gets a self-serve one.

Pair this with the lead scoring guide. Seniority says how senior the title reads; a decision-maker probability says whether that person can approve a purchase, and the two do not always agree.

Limits

Copy the template to try both formulas on your own contact export. The ICP fit guide uses the same HUNCHSCORE pattern on lead notes instead of titles.

Questions

Does company size change the seniority score?
Not unless it is in the text. The formula reads the title cell only. Add the headcount to the same cell, for example "Founder, 3-person startup" instead of "Founder" alone, and the score reflects what the sentence says.
Can I use my own department list instead of the one shown here?
Yes, any list up to 255 options, always ending in other. Match the list to how your own org chart splits, not a generic one.
Why return the seniority label instead of the number?
A label reads cleanly in a roster you scroll through by eye. Keep a second column with no "label" argument, the plain number, for sorting and averaging; the sentiment analysis guide uses the same pair for the same reason.
How many rows can one formula handle?
About 1,000 within Google's 30-second limit for a custom function. For a full CRM export, use Extensions › Hunch › Ask about selection, which writes values in chunks instead.
What if two people share the exact same title at different companies?
They can score differently, since the formula reads whatever else is in the cell along with the title, a company name or a team name you pasted in. A bare, identical title with no other context returns the same result for both.

Run it on your own column

Copy the template (or use the Excel add-in), paste your key, point the formula at your data. Every formula here has its Excel spelling underneath. 100 rows free to start, then $29 for 5,000 rows. Credits never expire and there is no subscription.

Reference: HUNCHPICK · HUNCHSCORE

More guides