Why Most ABM Programmes Fail, and the 18-Month Fix
Fullfunnel.io Co‑Founder Andrei Zinkevich unpacks why most ABM programmes quietly fail, how to merge demand gen and ABM into a single Allbound revenue engine, and why human‑first selling — not AI‑generated cadences — is still what wins complex B2B deals.
Topics
Account-based marketing has gone mainstream, yet many programmes still look like legacy lead‑gen wearing an ABM costume, built for dashboards, not buyers.
In this conversation, Andrei Zinkevich, Co‑Founder of Fullfunnel.io, walks through a 12–18 month roadmap for unifying ABM and demand generation into a single revenue operating model, grounded in real account progression rather than competing MQL and sourced‑by targets.
Across this interview, Zinkevich explains how mature teams quietly run dual tracks, keeping the legacy MQL machine alive while building proof for buyer‑needs‑first content, Allbound attribution, and cluster‑based targeting that reflects actual jobs‑to‑be‑done.
He shares concrete playbooks for pilot design, operating rhythms, and KPI protection that Martech and CX leaders can implement without blowing up their current dashboards or getting caught in boardroom politics.
Zinkevich has spent nearly two decades in B2B marketing, moving from enterprise roles at Kimberly‑Clark and Biosphere to founding his own consultancy and later co‑creating Fullfunnel.io, an ABM and demand‑gen firm focused on high‑ACV B2B tech and service companies with long, complex sales cycles.
Through his full‑funnel B2B marketing client work, he has become a leading voice on buyer‑centric, full‑funnel marketing that treats ABM as a cross‑functional operating system, not a campaign.
In the conversation that follows, he argues for “human‑first selling in the AI era,” where AI and martech handle 80% of the research and signal‑processing so humans can spend 100% of their energy on the 20% that wins deals: relationships, judgment and point‑of‑view.
He lays out governance mechanisms to end attribution wars, warns against over‑engineered ABM stacks that add friction for buyers, and offers an uncomfortably honest look at what it really takes for enterprises to move from volume‑driven lead‑gen to buyer‑led, Allbound revenue.
Mature teams are starting to integrate ABM and demand generation instead of treating them as competing budgets. What are the practical steps to move from parallel programmes to a single revenue operating model that Martech and CX leaders can realistically implement within 12–18 months?
Most teams try to merge ABM and demand gen by running them side by side and hoping they meet in the middle. They never do. You end up with two dashboards, two audience definitions, and two teams quietly competing for budget and credits.
A single revenue operating model takes 12 to 18 months. Here is the roadmap we create with clients.
Months 1–3: fix the foundation.
Three things have to happen first, or you just scale the current mess.
Get finance in the room and kill the sourced-by war. First-click vs last-click, marketing-sourced vs sales-sourced: in enterprise deals with 200+ touchpoints, it doesn’t make sense. Agree on one classification.
Any account where sales and marketing work together is an Allbound or ABM. The rep keeps 100% commission credit, so nobody’s motivation drops. Marketing gets judged on pipeline contribution and account progression, not a separate quota.
Change vertical targeting with cluster targeting. “All financial institutions” force generic messaging. Export your fastest and biggest closed-won deals from the last 12–24 months and group them by the same broken workflow and the same job-to-be-done.
Example: Ten companies that bought our ABM services because they need to do expansion among existing enterprise accounts make a cluster. The vertical doesn’t matter in this case. They have the same challenges and need the same outcomes.
Write real qualification criteria: Product-need signals, firmographic, disqualifiers, buying committee structure. That replaces the static sales wish list.
Months 4–6: run one small scope pilot.
Don’t try to roll ABM out company-wide. You’ll lose trust before you prove anything.
One cross-functional team: 1 marketer, 1 SDR, 1 AE. One cluster, one use case, 10 accounts. That’s the whole scope.
Protect the SDR’s KPIs for 90 days, in writing. If the rep is still measured on 100 dials a day, they revert to spraying volume by week two. Judge them on leading indicators instead: buying committees mapped, LinkedIn connection rates, thoughtful non-sales touches, real conversations started.
Set the operating rhythm. A weekly 30–60 minute pipeline sync where the team walks target accounts one by one and decides what each account needs next. A 15-minute daily standup to keep execution moving.
Months 7–12: systematise and document.
Once the pilot shows account penetration and sales book discovery calls with the strategic accounts, turn it into an operation.
Wire the CRM for account velocity. Replace the old lead lifecycle with stages that match how accounts actually progress:
- Cluster ICP: fits the criteria, doesn’t know you yet. Owned by demand gen and brand.
● Future pipeline: crossed your engagement threshold. Sales and marketing map the buying committee.
● Active focus: relationship built, need validated. AE and marketing run 1:1 activation.
Build the playbook vault. Turn the pilot’s steps into SOPs, scripts, quality benchmarks, and short training videos. Execution quality stops depending on who happens to be doing the work.
Use AI on the 80/20 rule. AI handles 80% of the research legwork: intent signals, earnings-call synthesis, and contact enrichment. Humans spend 100% of their time on the 20% that closes deals, the relationship and the 1:1 personalisation. Don’t let AI write the outreach.
Months 13–18: scale through a centre of excellence (COE).
Now you can extend across regions and teams.
Create a COE from your pilot team: a global ABM lead, a content strategist, an ops manager. They own the methodology, the CRM templates, and the quality bar.
Onboard 1–2 regional teams per quarter. Never all at once. You’ll under-support everyone and adoption dies.
Coach, don’t dictate. Local teams adapt the framework: analyse their GTM challenges, pick their clusters, adjust messaging for local competition, match local buyer channels, and set realistic
timelines. A low-awareness market might take 6–9 months to build a pipeline, even if your home pilot took 3 months.
Use 80/20 to structure your programme and keep innovating: 80% on proven always-on motions (reliable pipeline, thought leadership, customer webinars), 20% for the COE to test new channels and formats.
You’ve called for a shift from “SEO-first” content to “buyer needs first,” and from MQLs to engagement, pipeline and revenue as the KPIs. For CMOs still judged on traffic and leads, what does a serious transition plan look like — without blowing up their dashboards overnight?
Ditching MQLs and top-of-funnel traffic overnight is a fast way to get fired.
The average B2B marketing leader tenure is 16–18 months. Walk into a board meeting, announcing MQL playbook is dead, delete your lead-gen dashboards, and leadership assumes you’re hiding a weak pipeline. So you don’t blow anything up.
You run two tracks at once: keep the old MQL machine running while you quietly build the proof for the new motion.
Here is the 12–18 month plan.
Months 1–3: “add, don’t subtract.”
Keep every lead-gen and paid programme running. Your only job this quarter is to build tracking and gather evidence.
Layer blended attribution alongside the old reports. Keep the last click in HubSpot for leadership. In parallel, add a free-form “How did you hear about us?” field to every form, and have sales and CS ask it on every discovery and onboarding call.
Run a historical deal analysis. Export your 5 fastest and 5 largest deals from the last 12–24 months. Interview those buyers. Map the real journey. When you show your CEO a real customer with 240+ touchpoints over 9 months before they ever booked a call, last-click credit falls apart on its own.
Run the CFO quiz. Keep reporting traffic and MQLs, but sit your CFO in front of that journey map and ask: out of these dozens of social touches, podcast listens, and peer recommendations behind this $100k deal, which single one gets the credit?
That one question moves them from seeing marketing as a gumball machine to seeing a multi-touch system.
Months 4–6: the 90-day pilot.
Now that leadership is curious, run a tightly scoped pilot on the POS framework.
Keep the team lean. Half an FTE of a marketer, one motivated sales rep (not your loudest sceptic), a content resource, and an internal subject-matter expert.
Scope it down hard. One segment, one geography, one cluster. 10–15 supplier-unaware accounts (your Cluster ICP list) plus 30–40 showing light engagement (Future Pipeline).
Change the content for these accounts. Stop shipping keyword-stuffed SEO articles at them. Sit with your SME and build 3 pieces of real use-case content around their friction and jobs-to-be-done. Nurture with LinkedIn thought leadership, content collaborations (invite target buyers onto a podcast), and personalised content hubs.
Protect the pilot rep’s KPIs. Get sales leadership to agree, in writing, that this rep stops chasing 80 dials a day and gets measured on account progression and relationships built inside the buying committee.
Months 7–12: pivot the dashboard.
Once the pilot converts accounts at a higher rate, start moving your reporting over.
Introduce account velocity into board decks. Instead of “we generated 500 MQLs,” show the progression of your target account lists:
1. Cluster ICP: cold but qualified, building mental availability.
2. Future pipeline: hit your engagement threshold, buying committee being mapped.
3. Active focus: strong relationship, validated need, sales and marketing running 1:1 plans.
4. Sales opportunities: real deals co-created with sales.
5. Allbound revenue: closed-won.
6. Brand: mentions, speaking invites, organic traffic.
Adopt the Allbound agreement. Formally classify joint target accounts as Allbound with your CFO and CRO. Reps keep 100% commission credit. Marketing gets judged on the account-to-pipeline ratio (converting 5–15% of targeted accounts) and pipeline velocity.
Translate it into their language. Early buyer-needs content earns you access to the deal. Conversions come later. Frame it as: our 10-account pilot returned a 12% account-to-pipeline rate and $1.6M in qualified pipeline; matching that with volume outbound would have meant spamming 3,000 accounts.
Months 13–18: operationalise and scale.
Now you have the history, the pilot proof, and the CFO agreement.
Move to an 80/20 budget: 80% on proven always-on motions (cluster webinars, newsletter, sales enablement hubs), 20% on GTM experiments and new channels.
Build the playbook vault. Turn the pilot into SOPs, scripts, and training videos so you can onboard 1–2 new regions a quarter without quality dropping.
Form the COE from your pilot team. They coach other departments through adapting the framework to their markets.
You’ve argued for “human-first selling in the AI era” — that granular account research and POV-driven outreach still win the biggest deals. How do you see marketing and AI working together to give sales that human-level insight at scale, without turning everything into generic cadences?
The losing move is letting AI execute the outreach. That just floods inboxes with identical AI-written messages buyers have already trained themselves to ignore.
The model that works: AI does the enabling, humans do the selling. Run it on an 80/20 rule. AI handles 80% of the heavy, repetitive research legwork, so reps spend 100% of their energy on the 20% that wins deals, the relationship and the personalisation.
The 80%: where marketing and AI scale insight.
Instead of making reps dig through scattered data, marketing builds an insights engine that hands sales pre-analysed context.
Account research across pillars. AI digests earnings calls, press releases, hiring trends (a company throwing headcount at one workflow problem), and strategic initiatives, then produces a structured account briefing card.
Voice-of-customer analysis. Feed sales calls, discovery sessions, and CS onboarding transcripts into AI and pull out the exact jobs-to-be-done and the raw language buyers use. Reps drop those trigger phrases into their own messages so their POV matches the buyer’s reality.
Signal capture and prioritisation. Wire AI to watch first-party signals (pricing-page and case-study visits) and third-party intent, score accounts, and flag the ones moving from Cluster ICP to Active focus, so reps know exactly when and why to run a play.
Draft the raw material. AI writes first drafts of account love letters, LinkedIn posts, and content-hub structures so reps don’t stare at a blank page.
The 20%: what stays human, always.
Anything that needs judgment, credibility, or a real relationship stays manual. Automate these and your enterprise deals die.
Which accounts get 1:1 focus. AI ranks by data. Humans make the call, because they can read historical relationships, executive connections, and insider context AI can’t see.
The actual outreach. Messages, thoughtful comments on buyer posts, DMs, event hosting: written and sent by reps and execs.
The executive layer. Inviting buyers onto research projects or a podcast, hosting a small customer dinner. Enterprise buyers want human-to-human, and that can’t be automated.
Guardrails so you don’t scale a broken process.
The most common reason ABM programmes fail is scaling AI before the workflow works. Garbage in, garbage out, now at volume.
Master the process manually first. Run your 10-account pilot by hand before you write a single prompt. This is how you learn what “good” looks like.
Anchor every prompt to an SOP. A research prompt should specify the exact questions to answer, where AI looks, and where the validated data lands in the CRM.
Keep a human in the loop. AI hallucinates and pulls stale data, like a two-year-old initiative or a role that’s already filled. Reps check sources, timelines, and context before any email goes out.
We had a great case study with Backbase. Their CMO, Tim Rutten, cut black-box programmatic ad spend, built a custom AI signal engine to prioritise global accounts and do the superhuman research, and kept outreach, podcast hosting, and executive communities strictly human-led. Over 9 months, they flipped their target market from 60–70% cold to 60–70% actively aware and engaged.
In the AI era, your GTM strategy, your relationships, and your POV are the only things nobody can copy. Use AI to clear the operational runway so your reps can go build them.
Attribution in ABM is notoriously political: marketing, sales and CX fight over who “owns” deals when touchpoints span communities, events, ads, outbound and product. What governance mechanisms have you seen actually reduce these attribution wars?
To end the finger-pointing over who owns a deal, the best B2B teams drop marketing-sourced vs sales-sourced entirely. When an enterprise journey runs hundreds of touches across months, crediting one department or one last touch is operational fiction.
Here are five governance mechanisms that actually cool the attribution wars.
1. Classify revenue as Allbound.
Replace the sourced-by labels with two default CRM categories. Allbound: any deal from a target account where sales and marketing co-ran the plays. Marketing-influenced: a deal with verified marketing touches that wasn’t part of a joint programme.
The compensation rule is what makes it stick. The rep still gets 100% commission credit on Allbound deals. Marketing takes no commission and no separate quota; its success is tied to pipeline velocity and account progression. Now nobody has a reason to fight over credit.
2. Run the CFO quiz, then move to blended attribution.
You can’t change the rules without bringing finance along. Export 2–3 of your best recent wins and map every touch, from the first anonymous newsletter signup to the closing call.
Sit with your CFO and ask two questions:
- Out of these 80+ touchpoints, which single one gets 100% of the credit for this $100k deal?
- How do we calculate CAC if we pretend the last web visit was the only touch that mattered?
Once they see buying as an ecosystem, move official reporting to blended attribution built on three sources: self-reported (“How did you hear about us?” on forms and discovery calls), digital analytics (Dreamdata or HockeyStack for multi-touch), and customer interviews during onboarding to catch dark social and peer referrals.
3. Protect pilot KPIs.
Reps ignore joint playbooks because they’re chasing 80 dials a day while marketing chases MQL volume. That pressure creates the friction.
Get written agreement from sales leadership to protect the pilot reps’ KPIs for 90 days. Let them deprioritise outbound volume and judge them on joint leading indicators: buying committee members mapped and engaged, LinkedIn connection rates and real two-way conversations, and shared bridge activities like booking a target account into a strategy session or workflow audit.
4. Agree on success criteria before every campaign.
Before any event or cluster webinar, sales and marketing co-write and sign off on what success means. It kills the retrospective argument about whether it was worth the spend.
For a joint webinar, agree up front on how many priority accounts must register, how many buying committee members from those accounts must attend, and the exact follow-up playbook for both teams by engagement level.
5. Appoint one signal and data owner.
Siloed data is the root of account blindness: sales can’t see how marketing is engaging their accounts, and marketing can’t see sales’ conversations.
Give one GTM or RevOps owner clear accountability: centralise all first- and third-party signals into a single account card, qualify intent and route it daily or weekly to the right owner with full context (not blind automated sequences), and run the weekly pipeline review where the team walks live accounts one by one to plan joint next steps.
You’ve warned that many ABM playbooks are just “ads plus intent signals” bolted onto legacy demand waterfalls. What are the tell-tale signs that a company’s ABM stack is over-engineered for reporting but under-engineered for actual buyer experience?
When an ABM programme is built for the dashboard instead of the buyer, you’ve usually got a broken lead-gen programme wearing an ABM costume. The team bought expensive platforms to speed up a legacy demand-gen playbook instead of changing how they sell.
Here are the tell-tale signs.
1. Every signal gets treated as buying intent.
Internally, it looks great: dashboards lighting up with high intent scores and “engaged account” reports every time someone downloads an ebook or visits a blog.
For the buyer, it’s brutal. Someone in the “I just want to learn” phase gets hit with a 21-touch cold email and InMail sequence from an SDR pushing a 15-minute demo. You’ve turned curiosity into a sales handoff and trained the account to ignore you.
2. AI automation replaces a real point of view.
The CRM is full of automated funding, hiring, and technographic signals, and the team reports a high volume of “personalised” sequences at scale.
The buyer gets hollow templates: “Saw you just raised a Series B!” with zero value behind it. Half the time, the AI research is stale or hallucinated, referencing an initiative or a role from two years ago, and the rep looks amateur. Real personalisation means understanding the buying committee’s goals and building a POV worth their time.
3. Buying friction added for reporting convenience.
To keep attribution and CRM hygiene clean, marketing gates everything: long forms, hidden pricing, locked product info.
Modern buyers want to self-serve, see pricing, and try the product before they talk to anyone. Force them into a BANT call with a junior SDR who can’t answer a technical question, and they file you under an old-school supplier and drop out of the funnel.
4. Sales enablement is over-built, buyer enablement is missing.
Marketing shows off SEO traffic and whitepaper downloads as proof that content works.
But in a high-ACV deal, the buying committee has multiple stakeholders. Your champion is sold and then stranded trying to sell you internally. They don’t need another blog post. They need TCO breakdowns, implementation roadmaps, a custom business case, and security and procurement docs. When marketing doesn’t hand those over, the cycle drags and the deal stalls or loses to whoever made buying easier.
5. Siloed scoreboards and first/last-touch credit fights.
Marketing hits its MQL quota on its own dashboard while sales misses revenue and calls the leads garbage.
Because last-click can’t capture a non-linear 12-month journey full of peer chats, dark social, and private community mentions, leadership cuts the long-term brand budget. Ads, website, and outreach drift into separate universes, and the target account gets a disjointed, contradictory message.
ALSO READ: The Shift from ‘More Partners’ to ‘Fewer, Better Ones’
