Thomas Cornelius
LinkedIn snapshot
Post performance
Likes + comments per post, last 6 months
Top posts
- 1/29/2021textView →
After one year of intense software development work, I am so excited to share the merger of Hirebook and the leading OKR strategy consulting group PM2. Fortunate to now call Brett Knowles a partner at Brainware and Bre…
♥ 111💬 3↗ 0114 total - 3/16/2021mediaView →
It is not surprising, and amazing to witness how the constant focus on employees is at the core to build a workplace where clients, employees and partners thrive > our company Helpware has done that over the past years a…
♥ 103💬 8↗ 0111 total - 7/6/2025textView →
📈 Here’s how we operate graph8 — and how we keep every team member unblocked and producing daily. Maybe you find some nuggets in here for you out there pushing hard and building. We see you 🏗️ My message to our tea…
♥ 53💬 16↗ 069 total - 8mo agotextView →
We're spending close to $100K/month with AWS. Ran the numbers on moving to our own data center. Hetzner for compute, Cloudflare for storage and CDN. The math: 70% cost reduction. Not including LLM costs — that's a sep…
♥ 40💬 18↗ 058 total - 6mo agomediaView →
Today we launch something that shouldn’t exist. A single platform that handles B2B search, AI enrichment, campaign creation, multi-channel sequencing, personalized copy generation, and analytics. All self-serve. All API…
♥ 44💬 7↗ 051 total
Recent posts
All posts in the feed →- 2mo agotextView →
Every team has more intent data than ever and no more certainty about who is actually buying. Issue 03 of Programmable Revenue makes the case for the one test that sorts signal from noise. New issue every Tuesday. Re…
♥ 4💬 0↗ 04 total - 2mo agotextView →
A territory should tell a rep where to start. Too many teams give an SDR 3,000 accounts and call it coverage. The rep still has to guess who to call. The AE still has to decide which accounts deserve deeper work. The r…
♥ 22💬 2↗ 024 total - 2mo agotextView →
In Issue 04 of Programmable Revenue, we examine how to build territories around account potential, workload, and seller fit. The strongest evidence comes from a LinkedIn study of about 9,000 customer accounts. Reps usin…
♥ 13💬 0↗ 013 total - 2mo agotextView →
Last week we spent seven days asking one question: Which signal deserves a rep’s attention? 👉 Tuesday: We separated real buying signals from dashboard noise. 👉 Wednesday: We studied a lead score tested with a live sa…
♥ 3💬 0↗ 03 total - 2mo agotextView →
A lead-scoring model reached an AUC of 0.8161. That sounds impressive. In simple terms, it was good at ranking the leads most likely to convert. But a model can look good when tested on past data and still fail to impro…
♥ 8💬 1↗ 09 total - 2mo agotextView →
The accounts most likely to convert are often the ones that convert without you. If your intent signal targets them, it is taking credit for a deal it did not cause. This is the expensive mistake, and it hides because t…
♥ 8💬 1↗ 09 total - 2mo agotextView →
Most revenue teams ask a buying signal one question: Who is likely to buy? They should ask a second: Who is more likely to buy because we act? That difference matters. The first question helps you prioritize existing de…
♥ 7💬 0↗ 07 total - 2mo agomediaView →
A territory should tell a rep where to start. Too many teams give an SDR 3,000 accounts and call it coverage. The rep still has to guess who to call. The AE still has to decide which accounts deserve deeper work. The r…
♥ 13💬 0↗ 013 total - 2mo agotextView →
In Issue 04 of Programmable Revenue, we examine how to build territories around account potential, workload, and seller fit. The strongest evidence comes from a LinkedIn study of about 9,000 customer accounts. Reps usin…
♥ 11💬 0↗ 011 total - 2mo agomediaView →
Last week we spent seven days asking one question: Which signal deserves a rep’s attention? 👉 Tuesday: We separated real buying signals from dashboard noise. 👉 Wednesday: We studied a lead score tested with a live sa…
♥ 2💬 0↗ 02 total - 2mo agomediaView →
Most revenue teams ask a buying signal one question: Who is likely to buy? They should ask a second: Who is more likely to buy because we act? That difference matters. The first question helps you prioritize existing de…
♥ 6💬 0↗ 06 total - 2mo agomediaView →
The accounts most likely to convert are often the ones that convert without you. If your intent signal targets them, it is taking credit for a deal it did not cause. This is the expensive mistake, and it hides because t…
♥ 8💬 1↗ 09 total - 2mo agomediaView →
A lead-scoring model reached an AUC of 0.8161. That sounds impressive. In simple terms, it was good at ranking the leads most likely to convert. But a model can look good when tested on past data and still fail to impro…
♥ 8💬 1↗ 09 total - 2mo agotextView →
Every team has more intent data than ever and no more certainty about who is actually buying. Issue 03 of Programmable Revenue makes the case for the one test that sorts signal from noise. New issue every Tuesday. Re…
♥ 4💬 0↗ 04 total - 3mo agotextView →
Introducing Programmable Revenue: a weekly research read for people who build pipeline. "We need more activity" is almost always the wrong lever. Issue 01 does the math, with teardowns of how Kyle Norton, Stevie Case, an…
♥ 11💬 2↗ 013 total