The Signals Era: Why B2B Adtech Needs a Probability Mindset
Clicks lied. Cookies died. Now the future is probabilistic. Welcome to the Signals Era.
Section 1: From Determinism to Probabilistic B2B
For the past decade, B2B marketers were promised certainty: exact matches, perfect attribution, clear conversions. But the underlying infrastructure - cookies, deterministic CRM matches, tracking pixels - was never built to scale in enterprise buying cycles, nor withstand a world shifting rapidly toward privacy and fragmentation.
Briefly over the last 24 months or so, we lived in an ‘ID era’, where upwards of 400 ‘Unique ID’s built across a mix of probabilistic (IP/device/cohort) and deterministic (Hashed email, logins) signals trod a fine line in the grey zone of privacy. The objective was to rebuild the cookie in disguise, and it failed. The scaffolding is collapsing, with leading players facing significant and likely insurmountable privacy challenges. UID 2.0, TradeDesk’s flagship identity solution got flagged by the US Department of Health and Human services for its use in Healthcare advertising - which triggered a cascade of closer analysis. ID5, the European standard bearer pivoted in 2024 in response to regulatory pressure too, pivoting to a privacy-safe activation layer, essentially acknowledging that cross-domain identifiers are a shrinking currency. Google’s privacy sandbox and Safari’s dominancy in blocking third party cookies have tag-teamed to block cross domain identifiers in vast swathes. Arguably the final nail was Google retrenching from a 3rd party cookie cliff top moment.
The ID era was not a failure, but represented a bridge which bought time for the market to adapt.
That adaption is a a paradigm shift in how we model identity, intent, and activation.
That adaption is what the Adtech industry is calling the Signals Era - a phase where marketers must navigate incomplete data, fuzzy identity, and cross-channel fragmentation using probability, confidence scores, and multimodal inference. In short: B2B targeting is now an inference game.
And that’s a good thing - if you're ready.
Section 2: Why Probability Theory Is Now Your Core Growth Lever
We Were Never as Deterministic as We Pretended
Let’s be honest: even at the peak of the respective cookie and "ID Era," certainty was a myth.
Sure, hashed email (HEM) targeting felt deterministic. But what were we really measuring? A moment in time where a device, a user, and a behaviour converged - often with false confidence. A hashed email match doesn’t tell you who is holding the device, why they’re engaging, or whether it reflects genuine buying intent.
We’ve all seen the edge cases:
A senior buyer emails a link to a colleague; a demo form gets filled by an intern; a device is shared at home; an IP rotates across offices.
Even the cleanest deterministic system was built on probabilities disguised as certainties - then packaged as precision. The planner layer didn’t want grey, they wanted yes or no. Therefore the actors doing the most honest jobs would overly constrain their targeting to find the ‘yes’ opportunities, and probably 80% of the opportunities (or more) were in the ‘maybe’. Nobody wanted maybe, so potential pipeline gets ignored because the net is cast too tight.
Now, with channels like digital out-of-home (DOOH), audio, and CTV growing, the illusion fades entirely. These environments don’t auction impressions one by one. Identity is nascent and often missing, many of these new channels don’t match cookies or emails. And yet, they work. They build salience. They activate demand. They generate downstream lift. They move metrics that matter - but only if you're willing to accept slippage in exchange for coverage.
The modern marketer must trade surgical control for signal-rich exposure, leaning into probabilistic strategies with smart feedback loops.
And there's more change coming.
Programmatic is quietly shifting to batch
There are signs that even web-based programmatic could go batch-based in the near future:
Google’s Protected Audiences API (née FLEDGE) is explicitly batch-oriented - grouping users into cohorts, not individually targeting. Are Google preparing a rug-pull on the whole foundations of programmatic, much like they planned with cookies?
iOS privacy frameworks and aggregation techniques like SKAN are designed for deferred, aggregate-level insights, not impression-level logs.
Server-side bidding and SPO strategies from supply platforms are reducing the number of bid requests and emphasising contextual packages, not ID matches.
All of this points toward a future where per-impression decisioning is no longer the norm. The smartest B2B teams are preparing for that - not by resisting it, but by building adaptive, learning systems that model confidence over time.
Confidence Is the New Currency
In deterministic systems, every "yes" is a fact. In the Signals Era, we need to ask: how confident are we that this company, this intent signal, this account behaviour is meaningful?
The answer? Confidence scoring.
At FunnelFuel, we’re embracing Bayesian probability and signal weighting models to build a confidence layer across everything - from IP-based company and firmographic resolution to site behaviour inference to goal-path compression. Instead of discarding imperfect data, we model:
Multiple ‘wea'k’ signals (IP, device, session path, geo)
Strength modifiers (recency, intent layer, engagement scores, RFV velocity etc)
Decay curves and reinforcement from repeated behaviour
This isn’t just attribution. It’s about probabilistically ranking which accounts matter, and when. The outcome? Make media spend, outreach, and funnel strategy reflect likelihood, not false certainty.
Section 3: Cookieless Identity ≠ No Identity
The cookieless narrative often spirals into a doomsday story. But that’s a misread.
It’s been said that “B2B users don’t use cookies”. That’s false. B2B users absolutely do “use” cookies - they're the same people browsing in their personal and professional contexts. The problem isn’t cookie presence - it’s cookie relevance.
Cookies were always a fragile identity bridge in B2B - what they really amounted to is a link to one browser on one device. In their final evolution, some vendors layered on hashed email or login-based IDs to attempt cross-device stitching, but even those were probabilistic and often lacked transparency.
And in B2B, the entire foundation breaks down.
We’re not targeting individuals in linear journeys. We’re navigating buying committees that:
Span departments, job roles, and geographies
Involve an average of 6–10 stakeholders
Play out over months-long cycles with periods of intense activity followed by radio silence
Include offline and non-digital influences (sales calls, partner conversations, event moments)
A cookie tied to one browser - even a known user - offers almost no insight into this complex, multi-threaded journey.
So when cookies disappear, B2B marketers don’t lose precision. They lose an illusion.
What we gain, instead, is an opportunity: to rebuild identity around behaviours, patterns, and intent clusters that reflect groups of decision-makers in motion, not individuals in isolation.
What’s Changing — And What the Best Are Building To Win The Signal Era
This isn’t a superficial shift. It’s a deep infrastructural rewiring of how identity, intent, and targeting work. The winners in the Signals Era will be those who stop chasing IDs and instead start building signal-rich, confidence-scored journeys across time, devices, and channels.
a. IP Signals Are Evolving - But Under Pressure
IP targeting is not dead - it’s just maturing under duress.
Specialist vendors like IPFlow, Dealfront, Albacross, ZoomInfo, Clearbit/HubSpot have invested millions in IP-to-company resolution engines. These systems now go beyond CIDR matching to algorithmically uncloak ISP layers, bypass VPN fog, and triangulate traffic using probabilistic context like page type, session fingerprint, and time of day. They have tricks in their local that I’m not privy too, but their deep tracking of tens of thousands [combined] B2B sites across publishing and vendors is part of the source.
At FunnelFuel, our JetStream engine layers vendor signals with first-party analytics and session metadata to create a multi-source resolution score, increasing match integrity and rejecting false positives via outlier suppression.
But here’s the truth: IP alone is not enough - and it’s increasingly brittle. I still here of B2B specialists whose only specialism if bulk loading some IPs into a second rate DSP and targeting them on-mass. Its floored not least because;
Bidstream obfuscation (via privacy masking)
Middleware breakage (between SSPs, CDNs, and DSPs)
And a fundamental drop in IP reliability in programmatic pipes linked not least to privacy laws
This means you cannot just push a set of IPs into a DSP and expect accuracy. You must architect a signal integrity layer, combining:
Publisher-direct supply relationships (where IP is preserved)
Header enrichment techniques (e.g.
x-forwarded-for)And session stitching using other inputs like timestamps, referrers, or hashed email overlays
Anyone not actively engineering IP-resilience in their stack in 2025 is not a serious partner.
b. First-Party Intent Is the Foundation of the Modern Graph
What you can see directly is now your most strategic signal. If your analytics is not B2B ready and giving you a 4k UHD picture from a B2B perspective, then you have the wrong partner. A strong hint would be GA4. A topic for another day
Tools like mine at FunnelFuel, RudderStack, or Segment give you granular visibility into company-level patterns - even when IDs are stripped. When paired with:
CRM Events (e.g. form fills, demo booked)
Offline signals (e.g. event scans, email replies)
Hashed email re-matching
Referral patterns from known IPs or UTMs
You begin to construct a self-growing engagement graph - one that reflects not just visits, but patterns of repeat, escalating, and funnel-stage activity.
At FunnelFuel, we power this with a session-to-company bridge, then assign RFV-style scoring (Recency, Frequency, Value) over time to surface accounts surging with intent, even before they convert. These kind of feedback signals are the secret sauce, ensuring that the right accounts with the right signals are getting the right media dollar attention at the right moment.
Identity Graphs Go Probabilistic - and Cross-Channel
The ID graph isn’t gone - but it’s no longer deterministic. I no longer care for a straight ID graph personally; in B2B I care for the Account Signals Graph. I want IDs, but I also want any and all signals that I can throw into my AI/ML layers to make sense of the sheer chaos that an 11 month sales cycle throws up - including contextual fingerprints, channel specific specialist reporting (e.g. MAIDs that got close to my DooH ads and then visited the vendor website), time of day, geos, devices, HQ visits, first party website signals like security pages being ready to indicate that procurement is progressing to technical validation, purchase intent consumption, IDs, cookies, IPs, lat/long, and anything else that I can possibly lay my hands on. This is signals in motion, and big data doesn’t need that complete determinism when it can get scale in return.
Solutions like ID5, UID2, and Adform Fusion now lean into probabilistic resolution - using co-op login signals, device fingerprinting, and modelled cohorts to stitch together exposure paths across:
Display
CTV
Mobile in-app
Audio and DOOH
These graphs no longer claim to “know the user.” They instead model likelihood of exposure, frequency of brand contact, and shared signal lineage — allowing you to optimize creative, frequency, and budget at the account or group level, not user ID.
At FunnelFuel, we use a concert of graph partners and overlay it with our own internal session scoring - creating a dual-layer graph:
Programmatic exposure (probabilistic) + First-party analytics (behavioural) = Activation with confidence.
Contextual Becomes Trigger-Based - Not Just Targeting
Modern contextual isn’t just about what page a person is on right now, it’s about when, how often, and what comes next.
Here’s how the smartest B2B teams are using contextual today:
Using site type + time of day + referrer to infer intent clusters
Mapping channel-level IDs (e.g. SSP-specific segments, supply-path signatures) to inform funnel stage
Triggering omnichannel nurturing journeys: if someone hits a competitor comparison page at 10:30am, show a decision-support video on CTV that evening, then follow up with a retargeted podcast ad the next day
Contextual must have had 5 renaissances in the last 15 years, and invariably the industry complains that it ‘lacks scale’. By definition of a solution with that. level of discretion, of course it will, but it misses the point. These are tip of the pyramid signals, super hard to find and link to accounts in a world of LLM, masked browsing, work from home, fragmentation, cookieless browsers etc - so when you do, you use it for what it is - the ultimate signal amplifier
Contextual isn’t the fallback — it’s the new signal amplifier.
Scoring Models Unlock Precision in a Fuzzy World
With identifiers dissolving, the new precision is in scoring.
Top-tier B2B orgs now score accounts across dimensions like:
Recency-Frequency-Value (RFV): updated daily, across site + media + CRM
Content Velocity: how fast is the account consuming gated vs. ungated content
Intent Cluster Matching: is this account tracking against a known path to conversion?
Engagement to Activation Gaps: identifying where accounts stall, and nudging them forward with the right media type
Content type tracking: purchase intent is obvious, but I personally love looking at pages like API docs, security, SLA’s etc - if engineering and IT security is getting involved procurement is progressing to technical. This brings new faces into the buying committee - are your tactics adapting to these sorts of SIGNALs today?
This creates an account-level funnel state - not based on pageviews, but movement.
The playbook is no longer about stitching identity to a person. It’s about scoring identity to a confidence threshold, and activating at the account or buying group level. As new scores are reached on the leaderboard, the account progresses and new targeting, creative and attribution expectations should come with these shifts
Section 4: Omnichannel Is Not Optional Anymore
B2B buyers don’t live in a single channel. They oscillate between:
A podcast on their commute
A desktop session researching vendors
An out-of-home ad during a lunch break
A retargeted CTV ad at 8:00 PM
A whitepaper they get sent by a colleague
And yet, most B2B platforms treat programmatic media like it’s still banners and LinkedIn.
That’s the opportunity. Orchestration across CTV, OOH, banners, native, and audio - all stitched to the same account signal stream - is the new unlock. Not omnichannel for vanity. Omnichannel for probability stacking.
The more distinct signals we get per account across different surfaces, the higher our confidence and the smarter our optimisation logic.
Here’s what’s often missed in the cookieless / ID discourse: it’s not just adtech that’s changing - it’s how people use the internet.
And this shift is moving faster than anything we’ve seen in two decades.
Search intent is evaporating into LLMs
The old model of “search → visit → consume” is breaking down.
Large Language Models (LLMs) like ChatGPT, Perplexity, and Claude are intercepting the research phase - answering complex B2B questions before they ever hit a website or search engine.
These LLMs are now:
Shortcutting vendor evaluation
Replacing whitepaper downloads with summaries
Surfacing competitive matrices that used to require hours of browsing
Frankly, they are presenting vendor shortlists which in some cases won’t meet deeper inspection
LLMs are eating search, and this disproportionately affects B2B, where decision-makers are already time-poor and task-oriented.
The crumbs of intent that once littered the web - vendor review visits, analyst reports, keyword trails - are now increasingly locked behind chat interfaces and LLM infrastructure. Invisible. Untrackable. For now...
The Outcome? The content reservoir is draining
Organic search traffic is down 65% YoY for some B2B SaaS categories (source: SparkToro, 2024)
Google now sends < 25% of traffic it did in 2019 to third-party content (source: Similarweb)
The rise of zero-click results, AI overviews, and gated ecosystems is shrinking the open web
Vendor site stats are harder to find, but the odds are they are seeing less traffic then they once were too
In short: there’s less to track, less to target, and less to retarget - if you're playing the old game.
But B2B buyers are more digitally present than ever
And here’s the paradox:
Buying committees are larger than ever (Gartner: average 7.4+ stakeholders)
Stakeholders are more distributed, spanning countries, time zones, and job functions
Digital collaboration tools mean more of the buyer journey happens online, across devices, platforms, and channels
So while the surface has fragmented, the depth of opportunity is massive - if you can reach and interpret the signals.
Signal Driven Omnichannel Isn’t a Compromise - It’s a Power Move
B2B buyers don’t live in a single channel. They oscillate between:
A podcast on their commute
A desktop session researching vendors
An out-of-home ad during a lunch break
A retargeted CTV ad at 8:00 PM
A whitepaper they get sent by a colleague
And yet, most B2B platforms still treat programmatic media like it’s just banners and LinkedIn.
That’s the opportunity. Orchestration across CTV, OOH, display, native, audio, and email — all stitched together with signal-based scoring and identity probability — is how you win the Signals Era.
You don't need certainty. You need confidence, coverage, and orchestration.
And you need to build it now.
Section 5: What FunnelFuel Is Building for the Signals Era
Let’s make this real.
Here’s what our product team is actively building to thrive in this new paradigm:
JetStream Confidence Engine: A scoring layer that evaluates all incoming signals (IP, geo, time, content, source) and assigns confidence levels to account identity and funnel position.
High-Value Action Orchestration: Moving away from CTR, our analytics layer scores goals that correlate to revenue - form fills, calculator usage, demo interactions - and pipes those as weighted signals to DSPs based on our own proprietary engagement x intent matrix
Multi-Vendor Identity Enrichment: We’ve built a comprehensive enrichment engne, deduplicating and reconciling signals to create a cleaner account graph with match transparency to signals we can buy across on all channels
Omnichannel Signal Routing: Every signal in the system (Analytics, programmatic logs, form fills) feeds into our activation engine - allowing a single account journey to trigger the right next action, whether it’s an ABM banner, CTV spot, or site engagements.
Funnel-Aware Activation Logic: By clustering accounts by funnel stage — top, mid, bottom - we route different creative, bid strategies, and budget thresholds, all governed by signal strength and scoring velocity.
Final Thought: The Signals Era Favours the Probabilists
If you’re still chasing deterministic certainty, you’ll fall behind. The future of B2B marketing doesn’t belong to the platforms with the most data - it belongs to the teams that know how to weight it, score it, and orchestrate it across channels.
We built FunnelFuel for this shift. Because it’s not about fixing the cookie problem - it’s about winning with signals.
TL;DR: Why the Signals Era Is a Win for Product and Marketing Leaders Alike - How To Sell It To Your Org
The Signals Era isn’t a downgrade from the deterministic past - it’s a strategic upgrade built for how B2B actually works in 2025.
Here’s the good news:
You’re not limited by broken cookies or vanishing IDs anymore. You can now model real account behaviour across channels and time.
You can prioritise the right accounts faster. With confidence scoring, and engagement x intent scoring, you no longer guess who matters - you weight the right signals and move with purpose.
You can unify brand and demand. Contextual, CTV, OOH, and analytics can finally work together, not in silos, less restricted by the highly constricted yes/no methodology
You’re no longer chasing perfect identity. Instead, you’re orchestrating smart plays across buying committees, funnel stages, and attention surfaces.
This is not about losing precision. It’s about trading illusion for impact - and building systems that are more scalable, more flexible, and more aligned with how B2B decisions are really made.
The Signals Era isn’t scary - it’s a better blueprint.
It rewards smarter product thinking. It unlocks deeper marketing intelligence. And it gives you a right to play in the next decade of B2B growth.
👋 Want the B2B Signal Stack?
Hit 10 referrals and get The B2B Signal Stack: 7 Tools You Need to Replace Cookies & Win in 2025 - our exclusive strategy PDF.