Better Tech, Worse Data: Why B2B Marketing Databases Have Fallen Apart

Ok – so something of a rant here. I’ve been in data “hell” for a few weeks – ever since we contracted with Adapt, purchased some new data from them and then found that despite their validation claims, their data polluted our marketing data to the point we got flagged by our email provider – fabulous. Having dived into this, what have I found? Nothing good. We appear to have systemic poor standards and corruption in (what I see) as the entire data market. This would explain the rise of multi-source tools like Clay? Got any opinions on this? Let me know and in the meantime , read on.

We live in an age where the tools for B2B marketing and sales have never been sharper—AI-powered enrichment, intent-data platforms, predictive scoring engines, slick CRMs. You’d think this combo of tech prowess and data ambition would deliver richer, more accurate company and contact datasets than ever. Instead… something broke. And it’s not the tech. It’s the data.

The paradox of progress

Marketers today roll out campaign stacks that would have been sci-fi five years ago. But when you actually dig into the lists feeding those tools, the fundamentals are crumbling. Job titles are outdated. Companies have inflated or deflated headcounts. Contacts bounce or never existed. And worst of all, the vendors promising “95% accuracy” are still basing segmentation on the same shaky foundations.

Here are the facts:

• Nearly half of marketing data used for decision-making is incomplete, inaccurate or out-of-date.

• Industry benchmarks suggest B2B contact data decays by around 2.1% per month, or roughly 25% a year in many cases.

• A typical marketer believes 25% or more of their database is inaccurate; one survey found 60% don’t trust the health of their data.

So while the tools got smarter, the underlying records didn’t—and in many cases, they got worse.

Root cause #1: Self-reported firmographics (and the incentives to fib)

A lot of the trouble starts with the simplest assumptions: that the company size, industry, revenue and headcount numbers you use to filter accounts are actual facts. In reality, they’re often marketing narratives. Think of it:

• A company rounding up its headcount to look more credible.

• An executive maintaining a LinkedIn profile even after they’ve left.

• A directory scraping numbers without checking if those figures still hold.

When you use filters like “200–1,000 employees, $50–250M revenue, North America”, you assume those numbers are accurate. But many are based on self-reported or loosely gathered data. That means you’re already segmenting on guesswork. Add hybrid/remote workforce changes and more frequent restructures, and the error margin grows.

Root cause #2: Cheap offshore enrichment (lazy at scale)

Data vendors have had their own cost-pressure problem. To deliver millions of records cheaply, many rely on low-cost offshore teams tasked with scraping, verifying and populating fields. Under time pressure and volume quota, the temptation to shortcut looms large: guessing seniority, copying job titles verbatim, catching phone numbers that aren’t checked, submitting email addresses with minimal verification.

At small volumes, you might catch the errors. But at enterprise scale? You're drowning in junk. Thousands of contacts get flushed into your campaigns. Many bounce, some were never relevant, and your sales team wastes hours chasing ghosts.

Root cause #3: Dirty data circulating in an echo chamber

Think of the B2B data ecosystem as a big muddy pond. Even if one vendor tried to clean it up, the neighbors just pump the same water back through a different filter. Many vendors and enrichment platforms buy from each other, license the same public registries, and apply AI models trained on their own imperfect data. That means a wrong title or bad phone number doesn’t disappear — it gets recycled, re-labelled as “verified,” and sold as “premium.” Many practitioners report that all the big vendors feel like they’re pitching the same contacts under different branding.

What’s worse: your AI scorecard might be brilliant, but it's scoring data that’s fundamentally flawed. So yes—“smarter AI” makes faster decisions, but it doesn’t make the data any better.

The acceleration factors: Why it feels worse now

Several trends have made the decay problem worse:

• Job mobility is higher than ever. Emails and roles change frequently. Some analyses show business email bounce rates edging above 3.5% per month.

• Remote/hybrid work blurs the line between full-time employee, contractor, affiliate; the definition of “company headcount” is fuzzy.

• Budget and attention pressures mean companies invest in shiny campaign tech but leave cleaning and stewardship underfunded.

• AI hype drives more automation, which amplifies scale rather than accuracy. The faster you operate, the faster you mis-operate when the base data is broken.

• Everyone’s in denial, scared of losing their jobs, and pointing fingers at everyone and everything else for poor marketing yields.

  
  

Why this matters: Reach, relevance and reporting all take the hit

If you’re playing in the game of marketing-to-sales with bad data, the consequences ripple across everything:

• Reach suffers. Invalid emails, incorrect domains, wrong LinkedIn handles—you’re chasing shadows.

• Relevance suffers. You think you’re talking to an “IT Director at 500-employee SaaS firm,” but you’re actually reaching a consultant in a 50-person firm. Segmentation falls apart.

• Reporting suffers. Since your funnel inputs are dubious, forecasts, ROI, attribution—all become unreliable.

In other words: your big tech stack might deliver the dashboards, but the numbers they display are built on mud. Your sales team loses faith, marketing looks like it’s under-performing, and you spend more to get less.

A better approach: Fix the foundation instead of buying more tech

Here’s what you can do—and by “you,” I mean you, the sales/marketing/RevOps professional who’s sick of junk data.

1. Build smaller, fresher target lists

   Less is more. Focus on tighter, better-defined universes. Refresh them often. One small accurate list beats a million garbage records.

   
   

2. Treat third-party data as a hypothesis, not truth

   Use it as a starting point. Then layer on your own verification—behavioural signals, direct contact, validation calls, sales feedback.

   
   

3. Invest in data hygiene as a discipline

   Clean up bounce rates. De-duplicate records. Assign ownership—RevOps or a dedicated data steward.

   
   

4. Close the loop between sales and marketing

   If a contact is wrong, make the “mark bad data” action one click for sales. Feed it back to your data provider. Negotiate a deal with the vendor where you get replacement data, money back being as you're acting as their QA resource! Make accountability real.

   
   

5. Ask better questions of your vendors

   Don’t just ask “how many records do you have?” Ask “how do you source them?”, “how often do you verify?”, “how do you treat role changes, company dissolutions, remote shifts?” Maybe a waste of time - as we've said, the truth is in short supply generally these days, but specifically in this business it seems.

   
   

6. Avoid Offshore entanglements

   Hard to do these days, maybe impossible in this business, but if the price looks "too good to be true" it probably is!

Conclusion – It’s not a tech problem. It’s an honesty problem

The root issue in B2B marketing data isn’t that the tools are broken. It’s that we scaled the wrong behaviors: we assumed self-reported numbers were correct, we rewarded volume over accuracy, we recycled bad records and slapped “verified” labels on them. We automated the propagation of mediocrity.

If you want real improvement, it’s not about buying a new platform. It’s about changing your attitude toward data. Treat it as an asset. Treat cleaning as a process. Treat verification as non-negotiable. Until then, “data-driven marketing” is just a sophisticated way of being confidently wrong.

If you like this, I can pull together a slide deck too (you know you’ll want one) and a condensed version for a LinkedIn post.

AI and the Lighthouse: Why Some Navigate Further — and Others Run Aground

Rocks, monsters and storms - Can AI be the Beacon of the Friendly Lighthouse?

In modern life and business, we’re all convinced that we’re all captains (or leaders). Ever since Madonna asked us all to “strike a pose,” we’ve been conditioned over the last 30 years to picture ourselves steady at the helm — confident hands on the wheel, charting the course. We’ve got it all figured out and cannot admit to any doubt. Then along comes AI: maybe a lighthouse in the fog?

AI shines a light on everything — the rocks, the routes, and the shortcuts. The recent MIT research, Economist article and observations on “dirty data”, serve to shine their own lights on existing (and sometimes perpetual) problems – illusions of competence, status quo bias,  and old baggage (which as U2 point out – “we can’t leave behind.”). These have inherently little to do with AI, but suddenly, the sea feels different. The question is: does the light of AI make us better navigators… or just more aware of how lost we really are?

Two Views of the Same Horizon

When it comes to AI’s impact on performance, two competing schools of thought have emerged.

One view, championed by Ethan Mollick, author of Co-Intelligence: Living and Working with AI, argues that AI acts as a great equalizer, a democratizer of knowledge - the light that helps the average navigator find their bearings. Mollick argues that AI gives the biggest boost to mid/lower-level performers, helping them think and problem-solve with more structure and insight. In his words (not mine), “it turns poor performers into good performers.”

The other view, recently published in The Wall Street Journal by Matthew Call of Texas A&M, suggests that the opposite is true — that AI widens the gap between top performers and everyone else. A bit like the way the “asset appreciating” economy of the last 15 years has favored those who already have/had assets to appreciate, opening up a bigger gap between the “haves and the have nots.” Call’s research finds that high performers are the ones who adapt first, learn faster, and use their expertise to ask more nuanced questions — seeing the increasingly important details that others miss.

So which is it? Is AI the lighthouse that helps everyone steer straight — or a beam that blinds the unprepared?

When Confidence Becomes Complacency

The answer, as it turns out, depends less on who’s holding the wheel, and more on what you want the wheel to do! . How so?

Top performers — the experienced captains — tend to embrace AI with curiosity and intent. They experiment, and use the light to read the waves ahead. They know when to trust their instruments and when to trust their instincts. They might also suffer from the complacent reaction of “I already knew that.”

Meanwhile, many others — often just as capable but less sure — either wait for “official guidance” or assume the light will do the steering for them. In other words, they confuse illumination with navigation.

This is where the modern workplace illusion creeps in: most people think they’re already captains. We’ve been conditioned to believe we’re all high performers. But AI doesn’t reward self-belief — it rewards skill, adaptability, and humility. It doesn’t care who you think you are. It cares how you think about using it.

The Lighthouse Exposes the Rocks — It Doesn’t Move Them

AI reveals more than it replaces. In sales and marketing, for example, AI can cut through the bias and drudgery of account planning. It gathers data, surfaces insights, and highlights connections that human teams miss. Tools like Shadow act as specialized co-pilots, giving teams greater “intellectual horsepower” by quickly doing some heavy lifting on research, analysis and suggesting obscure but powerful approaches.

But that same light can also expose weaknesses. AI doesn’t fix a poor strategy — it just makes it more visible. It doesn’t stop a team from being biased — it simply shows the bias in higher resolution and draws our attention to it. And if people overestimate their own capability, they’ll still end up sailing toward the rocks, they’ll just do it faster. The equalizer and the divider, as it turns out, are the same beam of light.

How Leaders Keep the Fleet Afloat

So how do you make sure your people don’t get dazzled — or shipwrecked — by the glow of AI? Call’s research offers three good navigational principles:

• Create safe harbors for experimentation. Give teams time and space to explore AI tools without fear of getting it wrong. You can’t learn to navigate by staring at a manual, as sailors know, you have to get wet.

• Share the maps. Capture and circulate effective learnings so one person’s discovery becomes everyone’s advantage.

• Redefine what “good” looks like. Update evaluation systems so AI-assisted work is recognized fairly, not dismissed as “cheating” or credited only to the usual stars.

  
  

In short, AI is less of a compass, and more of an environmental shift — one that requires new habits, new metrics, and a new humility about what it means to be “good” at your job.

Conclusion: The Light Shows What Was Always There

In this regard AI is no different from other technology waves. It’s not THAT you use it, it’s HOW you use. It’s application over availability.

For some, that’s liberating — the light helps them see paths they never knew existed. For others, it’s unsettling — the fog was hiding more than just the shore.

The truth is, AI isn’t inherently an equalizer or a divider. It’s a mirror held up to how your organization learns, shares, and adapts. Lighthouses don’t steer ships. People do. And AI won’t futureproof or propel your organization in and of itself. By applying it courageously and thoughtfully, you will.

Cold Calls, Warm AIs: Business Development is Dead...Long Live Business Development!

The Death of the SDR?

In the last few months, something quietly remarkable happened in B2B sales. A few of the companies that built their entire businesses on enabling SDR teams — names like Salesloft and Outreach — continued laying off their own SDRs. If that sounds ironic, it’s because it is. It’s like McDonald’s announcing the end of the burger. These are the companies that industrialized prospecting — fueled by sequences, cadences, and dashboards. Now they’re admitting what many of us believed (but got shouted down for saying) – that all their technology enabled people to do was to continue to commit marketing & sales crimes and misdemeanors, just in greater numbers!

I know, I did it

In ’93 (1993, NOT 1893, thanks!) we started a company that was built to do this – outbound B2B lead gen for tech companies, and it was new back then and fabulous! Other things (like email) emerged, but what really put the hammer to it was …you guessed it, everybody’s friend…Offshore. What happened next? The application of technology to compete with cheap labor offshore - the same as is happening now in the manufacturing space. SalesLoft (and the like) convinced companies to use their tech so their SDR’s could make 200 daily calls as opposed to 100 calls. As “lead rates” fall (as people become fatigued with the increasing volume of calls they receive as well as the declining quality ()people who can’t speak English)), then the “promise” of a Salesloft and many others was to enable you to keep cramming bigger numbers at the top. The math was misleadingly simple – when lead rates half, simply make twice as many calls! A race to the bottom.

The False Promise of the SDR Machine

The Sales Development model was, for a while, the closest thing to a religion in SaaS. Hire a small army of young, ambitious SDRs. Arm them with sales tech, email automation, and phone-dialing software. Then measure them — mercilessly — on volume.

More calls. More emails. More sequences. Because, as the logic went, “sales is a numbers game.” Except… the math stopped working. (never worked.) Buyers installed spam filters, ignored cold calls, and tuned out generic outreach. SDRs became noise machines — activity-rich, outcome-poor.

The tools that promised efficiency delivered fatigue — both for the prospects and the SDRs themselves. The model started to resemble a treadmill: lots of motion, very little progress.

Even worse, the function began cannibalizing trust. The more automation crept in, the more robotic outreach became — until prospects stopped answering altogether. And now, the irony is complete: the companies that powered that model are abandoning it internally.

Go to the Pod AND the video!

AI Blows Up the Math

The arrival of AI has been like pulling the plug on a faulty machine. Suddenly, sellers can do what used to take entire teams weeks. And they can do it better, faster, and with infinitely more context and relevance.

In other words, AI doesn’t just make prospecting faster — it makes random prospecting obsolete. It actually delivers real personalization, at (more) scale. In other words you can now achieve something that was a contradiction!

The Return of the Self-Sufficient AE

The pendulum has swung back to where it arguably always should have been — to the Account Executive who owns both creation and conversion. But this isn’t the “lone wolf” AE of the past. This is an AI-augmented AE — a seller who uses technology to exponentially increase their capabilities. Want a visual? Think Sigourney Weaver in Aliens, Tom Cruise in Edge of Tomorrow and Matt Damon in Elysium.

The emerging AE doesn't rely on inbound lists or SDR hand-offs. They use AI to surface insights, build narratives, and leverage context. They sound human because they are human — supported by tools that understand details and nuance. And in crowded, loud, homogeneous markets its ALL about details and nuance.

In this model, business development doesn’t disappear. It becomes part of the AE’s rhythm — a seamless blend of discovery, preparation, and engagement. The difference is that machines do the mining; humans do the meaning.

What Business Development Really Means Now

The phrase “business development” came to mean volume. An arbitrary number of opens, click, dials, touches led to “leads” and meetings booked. Now it’s about value discovery — finding where your offer intersects with a prospect’s opportunity or threat.

It’s not about who can contact more people. It’s about who can contextualize faster.

And that’s exactly where well applied AI excels.

AI can digest a target account's information, find and connect synergies and contextualize this across the landscape. That's not so much personalization as it is relevance. We've moved from the information age to the interpretation age. It doesn’t mean no one picks up the phone anymore. It just means they do it at the right time, for the right reason, and with something righteouse to say!

The ADR Era

So, if the SDR is fading, what replaces it? Maybe it’s something new: the ADR — AI (enabled) Development Representative.

An AI-driven system that:

• Researches markets and accounts.

• Flags opportunities based on synergies and timing.

• Drafts smart, relevant outreach communications.

• Hands the human a complete, contextualized picture of an opportunity.

  
  

The ADR doesn’t send 1,000 emails. It identifies the five conversations worth having, and gets the AE prepared faster and smarter. And in that world, “business development” doesn’t vanish — it finally grows up.

Full Circle

Ironically, this evolution brings B2B sales back to where it began. real people, having real conversations, about real business problems. When I started in this business sellers were responsible for their own prospecting right way through to their own closing – from first call to final close. The difference now is that the preparation is automated. AI doesn’t replace the human; it redeems them from bad process.

Maybe that’s the final twist in this long story. The SDR era didn’t die because of technology. It died because of its own lack of humanity. And AI — the very thing people feared would make sales less human — might be what makes sales more human…again.