About
I help software companies decide where to place their bets — then execute them, with…
Articles by Rajat
-
A Dedication ...
A Dedication ...
The last time I felt pride like this, Bill Coleman was handing me a magazine. It was the September 18, 2006 issue of…
21 Comments -
Is your data a moat, or just a feature?
Is your data a moat, or just a feature?
It was a loud week. SpaceX buys Cursor for $60 billion, Salesforce buys Fin for $3.
1 Comment -
Your intent data isn't the problem. Your interpretation of it is.
Your intent data isn't the problem. Your interpretation of it is.
Everyone in B2B has intent data now. Most teams I talk to are not getting any better at using it.
Activity
2K followers
-
Rajat Bhatnagar shared thisProud Dad Moment We attended a beautiful White Coat ceremony at the New York Academy of Medicine this afternoon for my son who started Medical School at Touro University this week. He was 1 of 135 lucky applicants from over 11,000 who applied in this cycle that made the cut. Touro, which ranks right behind the University of California system in producing medical professionals accounts for 4% of all new doctors in the United States each year. One fine day more than 20 years ago, as a precocious 4-year old, he began proclaiming to anyone who would listen that he was going to be a Doctor when he grew up. We all laughed but learned soon enough to not doubt his resolve. Today, he completed the first and most important milestone in that very long journey. Rohan Bhatnagar, your family (Ujjual Bhatnagar, Jai Bhatnagar and I) are so proud of you. Go be the great physician you’ve dreamt of being, that we know you’re meant to be!
-
Rajat Bhatnagar shared thisA Dedication ... The last time I felt pride like this, Bill Coleman handed me a signed magazine. It was the September 2006 Forbes cover — "The New Barbarians" — the first time in his life he'd made the front of a magazine like that. I was point on the customer deals the story was built on: Pfizer, Bank of America, NASD, U.S. Joint Forces Command. Those were my accounts, and my customers were the ones Forbes interviewed. He knew whose work put him there. He signed it: "Rajat, You are the greatest! All the best! Bill Coleman." I still have it where I can see it. This week I published a book, "The Campaign." It's dedicated to Bill (1947–2020), the founding CEO of BEA Systems — the "B" in BEA — and the mentor who taught me most of what I know about this industry. He kept a $250M philanthropic pledge after the dot-com crash cut his fortune from roughly $900M to $50M, when walking away would have been easy. He answered every cold email from anyone who'd worked at BEA, because he never forgot what they'd built. He gave his time to causes far bigger than any company he ran. Integrity. Gratitude. Service. If you knew Bill, I'd love your story in the comments — and I'd like to find the others he mentored while we still can. #Leadership #Mentorship #SiliconValley
-
Rajat Bhatnagar shared thisAnyone who's known me longer than five minutes has heard me bring up The Medici Effect. Johansson's argument stuck years ago and never left: the good ideas don't come from going deeper into your own field, they come from two fields you'd never put together bumping into each other. I've said that for years. Didn't expect to walk straight into one. A Roman election in 64 BC and a modern B2B deal, wired more or less the same. Quintus Cicero wrote his brother a handbook for winning that election. Practical, almost blunt: work out who actually holds sway, lock in the people who already owe you, give them something to hope for, don't walk into the Forum by yourself. The brother won the seat. Read it as a marketer and it stops feeling like history. The buying committee is an electorate. Your champion is whoever's arguing for you in a room you'll never sit in. The skeptic, the competitor, the procurement lead who can kill a deal without taking the meeting, that's just the opposition, and it's older than software. So I wrote it down. The Campaign: Cicero's manual as 22 principles for B2B, three case studies, and a chapter of AI prompts so you can actually use it instead of nodding along. What stays with me is Johansson's point. There was no new tool involved. The whole edge came from reading a 2,000-year-old book sideways, as marketing instead of history. A few pages below: the cover, the contents, who it's for, and the first chapter. Has anything collided like that in your work? Two things with no business going together that somehow did. #B2BMarketing #GoToMarket #TheMediciEffect
-
Rajat Bhatnagar shared thisAI Consolidation & The Proprietary Data Moat Everyone's reading this week's AI deals as a consolidation wave. $60B for Cursor, $3.6B for Fin, both in a matter of days. I think the deal that matters more closed a month ago and barely registered: Publicis buying LiveRamp for $2.5B. That one wasn't about software, or talent, or growth. It was about owning an identity asset that 25,000 publishers already run through. The kind of thing you can't rebuild no matter how good your models get. To me that's the real signal under the noise. As AI eats the application layer, the value is moving to proprietary data that can't be copied. Some companies are about to get repriced upward for it. A lot of others are going to find out that "AI-powered" was never the moat they thought it was. I wrote up where I think the line actually falls 👇 https://lnkd.in/g-N3gx9j #MergersAndAcquisitions #DataStrategy #CorporateDevelopment #EnterpriseAI
-
Rajat Bhatnagar shared thisAs AI eats the application layer, the value is moving to proprietary data that can't be copied. Some companies are about to get repriced upward for it. A lot of others are going to find out that "AI-powered" was never the moat they thought it was.
-
Rajat Bhatnagar posted thisYour intent data isn't the problem. Your interpretation of it is. Three accounts spiked on the same topic this week. One has no install. One already buys from you. One runs a competitor. Same signal. The right play for each is completely different — and at most companies, all three get the same email. Here's what I think is going on. Intent data isn't the bottleneck anymore. We have plenty. What we don't have is the discipline to ask, for each surge: what does this mean given what the account already owns, and is it worth a campaign dollar? I wrote a longer piece on this. A few things I keep coming back to: The motion math. Renewals close 80–90% of the time. Expansion 60–70%. New logos 20–40%. Competitive takeouts 5–15%, and the cycles run 2–3× longer than net-new. Put those next to each other and the spend priority is hard to argue with. There are five motions, not three. Most plans I see name Whitespace, Expansion, and Displacement. They quietly skip Churn Prevention and Dark Funnel — which is also why those two keep eating budget without anyone noticing. The number I can't get out of my head: a churn signal you miss now costs 4 to 10× what it would cost to win that same revenue back as a new logo a year and a half from now. Most companies are still spending more on new logo acquisition than on actually reading what their own customers are doing. TAM, SAM, SOM, ICP all do different jobs and most teams use one of them for everything. The part that gets skipped: ICP should be bounding the other three. If your SAM doesn't have an ICP filter on it, you're working off a fantasy number. Full piece in the first comment. Would genuinely like to know what other people are seeing — especially anyone running a renewal motion against a determined competitive takeout campaign.
-
Rajat Bhatnagar shared thisThree accounts surge on the same topic this week. One has no install. One already buys from you. One runs a competitor. Same signal trace. Three completely different campaigns. And in most B2B orgs, all three get the same generic email. I wrote a piece on why this keeps happening, what the actual economics of your GTM motions look like when you do the math, and the one piece of the TAM/SAM/SOM/ICP conversation almost everyone skips. Worth a read if your team is sitting on more intent data than it's getting value out of.
-
Rajat Bhatnagar posted thisPART II Is ARR the right way to value a data company? So, the genuine question, not a rhetorical one: if the enduring asset is a unique, trusted, continuously refreshed body of data with a persistent identity, is ARR really the right way to express what we're worth? ARR measures the rhythm of the contracts. It says almost nothing about the depth, uniqueness, or half-life of the underlying data, which is the thing that will still matter after this model generation and the next one are obsolete. I'll state my conviction plainly, then I'd love to be argued with: AI runs the same arc every important technology runs. Boom, bust, build. We are deep in the boom, and a lot of today's exuberance is frankly irrational, some of it engineered to pump valuations so the loudest raiser survives. The Internet did this in 2000-2001. The crash didn't kill the Internet. It cleared the field, and the companies that had built something real came out the other side much bigger and much stronger. AI's bust, when it comes, won't be the end of anything. It'll be the moment the market stops paying for the story and starts paying for the substance. And the substance comes back to data. So, over to you, especially if you're building in this space: 1. If you run a data-centric business, what metric actually captures the worth of your data asset, the way ARR captures the rhythm of subscriptions? Does one even exist yet? 2. Does the data moat really strengthen as models improve, or am I telling myself a comforting story? 3. Boom, bust, build, then the data businesses inherit the build phase. Right read, or is this time different in a way I'm underweighting? I don't think ARR is finished. It's had a remarkable run and probably has miles left in the tread. But for companies like ours, whose value lives in the data layer and not the code layer, I suspect it was always describing the surface rather than the foundation. The companies still standing in fifty years won't be the ones who optimized the metric. They'll be the ones who built the thing the metric was always a clumsy proxy for. Curious whether you think I've got that right. #AI #DataStrategy #SaaS #Valuation #PrivateEquity
-
Rajat Bhatnagar posted thisPART I Is ARR the right way to value a data company? I spent a long graduation weekend half-listening to the usual chatter (Nvidia, the Iran war, somebody's PE exit) through the one filter I can't switch off: what does any of this mean for the company my team and I are actually building? By Tuesday a post making the rounds had sharpened the question better than I had all weekend. The claim: ARR isn't a business model. It's a 27-year anomaly, and it's ending. The history is hard to argue with. Software used to be a thing you bought once and owned. Salesforce turned it into a subscription, IPO'd in 2004 with barely any earnings, and the market reached for a new ruler: Annual Recurring Revenue. By 2018 ARR was close to scripture. By 2021 the best companies were trading near 18x revenue and the wildest private rounds went far past that. "100x ARR" became a punchline. Then 2026 happened. Software's forward P/E slipped below the S&P 500 for the first time anyone can remember. Public SaaS multiples fell back to 6-7x, roughly where they were in 2015. The market quietly started pricing on cash (EV/EBITDA, Rule of 40, net revenue retention) instead of contracted intent. Here's the part I keep chewing on, though. ARR didn't fail. It succeeded too well. For most of that run it was a perfectly good proxy for a real thing: a durable, un-replicable SOFTWARE moat. The trouble is that the software moat is exactly what AI is eroding, because AI makes code cheaper to produce by the month. So if your defensibility lives in the code, the obituary is aimed right at you. But my company is a data company. The code was never the point. The data is the point. And a deep, clean, well-governed, hard-won dataset doesn't get cheaper to copy when models get better. It gets MORE valuable, because every capable model in the world now needs something true to stand on. There's no good AI without good data. That's not a slogan to me. It's the whole thesis. I think about Dun & Bradstreet here. Founded in 1841. Still a data business nearly two centuries later. Its D-U-N-S Number (1963) gave companies a persistent identity the entire commercial world quietly came to depend on. Empires of software have risen and fallen in the time D&B has simply kept being the place you go to know who a company really is. (And worth noting for anyone who assumes PE means short-term: D&B today is privately held, PE-owned. Patient ownership is often exactly how these data franchises got the runway to compound across generations. My own company is 15 and PE-backed, and I've come to see that as continuity, not tension.)
Rajat Bhatnagar commented on a post
2d
Stacy McMaster may I suggest that you start with black raspberry ice cream whey protein. It’s guaranteed to fly off the shelves!
-
Rajat Bhatnagar liked thisRajat Bhatnagar liked thisThrowback to when Outlook 2026 took home gold at the NYX Awards. The win landed in the B2B Branded Content Campaign category, which honors creativity, performance, and innovation across industries, and we could not be prouder of the team behind it. What earned the trophy: 📊 Exclusive insights from 2,300+ go-to-market experts 🤖 AI-powered personalized intelligence 🎯 Role-specific frameworks matched by industry Explore the award-winning Outlook 2026 hub: https://lnkd.in/dujtbyDE #TBT #NYXAwards #Outlook2026 #B2BMarketing #DemandGeneration
-
Rajat Bhatnagar liked thisRajat Bhatnagar liked thisDoes anyone know if black raspberry ice cream is hiring? I’d like to submit myself for consideration. It’s the superior fruit inspired ice cream. No disrespect to strawberry. Strawberry frappes are a staple in my life. But black raspberry? Heaven. It’s so slept on but with a little rebranding and maybe some targeted ads I believe I could bring it to the masses. I understand this varies regionally but according to multiple studies, black raspberry ice cream is not in the top 10 in the US and depending on the survey, doesn’t even break the top 20! Who can give me first party “searching for ice cream” intent? Anyway, I don’t want summer to end. #saturdayfun #icecream #intent #sundaes #signals
-
Rajat Bhatnagar liked thisThank you to Technical.ly for recognizing me as a 2026 RealLIST Innovator in DC! It is great to be recognized for the work I've helped contribute to both at GTM Fabric and The Upskilling Labs. This is only possible because I get to work with amazing people every day, who encourage me to be my best self and drive towards the things I'm passionate about doing. And it's even more exciting to be named alongside other members of the The Upskilling Labs community!Rajat Bhatnagar liked thisTechnical.ly named its 2026 RealLIST Innovators. Four people who build with us made the list. EChO (Eleanor Ory), Ph.D. — pilot alumni, founder of RideFlow DC Ashwin Jaiprakash — strategic advisor, VP of AI, Product, Engineering, and Growth at GTM Fabric Shawn Gregoire — mentor, founder of Capstone Tech Nicholas Wagner — mentor, CEO of Learning Journey AI Full list: https://lnkd.in/evZRapEc Thank you Kaela Roeder and Danya Henninger of Technical.ly for the reporting. Thank you DC Public Library — Chelsea Kirkland — and Levy Strategic Design — Ann Marie Guzzi and Brendan Whitaker — for the partnership that made this work possible. #DCTech #AI #Upskilling #Innovation
-
Rajat Bhatnagar liked thisRajat Bhatnagar liked thisDoes anyone else ever randomly think about the incredible people they’ve met through their career and how they have made their life better? Just me? Yeah yeah I know “work isn’t about making friends” but some of the best people I’ve ever met, and many of my closest pals and confidantes all came into my life through professional situations. Anyway, I just had a moment where I realize how genuinely lucky I am to know and have learned from so many amazing people. Gratitude is everything. If you caught the reference in this post, we can be friends. #thankyouforbeingafriend #mentorship #friendsareimportant #leadership #connections #gratitude
-
Rajat Bhatnagar liked thisRajat Bhatnagar liked thisYour ICP is fine. The issue is your list-building process. Your target account list has a ghost problem: contacts who left, companies that got acquired, accounts still scoring high because nobody touched that CRM field since 2022. Meanwhile, your real best-fit accounts—those running the right tech stack and spending in the categories that signal "buy from us"—never appear. Technographic and IT spend signals don’t live in standard CRM fields, so they never make it into your scoring. That’s the actual gap: not your ICP, but your CRM’s ability to act on it. Here’s a number that stuck with us: Airbase’s identified SAM in Salesforce grew 80% overnight after they layered in signal-backed data. Same company. Same CRM. Suddenly an 80% larger addressable market. The accounts were always there. The CRM simply couldn’t see them. If you’re a RevOps leader who suspects your pipeline is missing your best accounts, we wrote a starter kit on exactly this gap and how to close it. Link in comment below.
Experience & Education
Languages
-
English
Native or bilingual proficiency
-
Hindi
Native or bilingual proficiency
Recommendations received
2 people have recommended Rajat
Join now to viewOther similar profiles
Explore more posts
-
AmyGB.ai
Hyperautomation brings AI, automation, and decisioning together to handle document-heavy tasks end to end. AmyGB applies this across extraction, verification, enrichment, and routing so teams avoid repetitive review cycles. Each document moves from intake to structured output with consistent accuracy. This reduces operational load and strengthens onboarding and compliance workflows. A practical path to faster decisions and cleaner data. #Hyperautomation #IntelligentDocumentProcessing #EnterpriseAI #WordoftheDay
-
Harry Ault
Think of a world without server virtualization…. Dinosaur days of one app, one server… Ironically, this is how GPUs are deployed around the world - one model, one GPU. Extremely inefficient unless you are at massive scale. Enter Sambanova’s unique model bundling capability - run multiple models on one RDU. The future, efficiently delivered in production TODAY! Check out the end to end dev experience here, game changer in blazing fast inference for agentic apps. 😎👍
1 Comment
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content