The short answer
• B2B marketing teams use about half the martech they pay for. Gartner put utilisation at 49% in 2025.
• Forrester principal analyst Rani Salehi lists "technology is an enabler" as one of ten rules for go-to-market planning. Disconnected systems are her example of it going wrong.
• Most of the stack still does its job. The marketing automation platform is where it breaks because it was built for one lead and B2B now buys in groups.
• The fix is fewer layers that actually connect. Audit before each renewal, learn how your buyers buy, rebuild scoring, routing and decisioning and start with your own data.
• Clay brings data and signals together. Phave replaces the rules-based marketing automation platform. Both usually sit on the CRM you already have.
• Retire two tools for every one you add.
Most finance teams see martech as a cost centre, and in 2026 they had the numbers to prove it. Gartner says B2B teams use 49% of the martech they pay for. Zylo says 46% of licences go unused. A CFO looking at those figures sees software spend with no clear link to revenue.
That changes in 2027. The budget for next year is going to tools that can show the pipeline they create. In ABM that means tools that find the buying group, decide who gets the next touch and hand sales a qualified account rather than one form fill. Clay and Phave are built for that job. Shelfware gets measured in licences. ABM automation gets measured in opportunities.
The teams that make that switch in 2027 will spend less on software and be able to show the board what it earned.
Ask marketing leaders about their stack and you get surprisingly honest answers. Managing director after managing director tells me parts of it aren't being used. A couple have admitted they don't know what some of their tools do.
And the same problem keeps coming up. It's too hard to get a new tool approved internally so implementation keeps slipping down the list. That happens even when the current set-up is slowing everyone down.
The knock-on effect comes up almost as often. Because the stack is fragmented they don't trust their CRM data anymore.
Meanwhile the stock market has its own word for software: SaaSpocalypse. In the weeks after 28 January 2026 roughly $1 trillion came off the S&P 500 software index. The bet was that AI agents would replace software user seats (SoftwareSeni).
I think the market priced the wrong problem. Marketing teams already own far more software than they use.
So how much of the martech stack do B2B teams functionally use?
• 49% of martech capabilities were in use in 2025 according to Gartner. It was 33% in 2023 (Gartner).
• 305 SaaS apps sit in the average company's portfolio in 2026 and about nine more get added each month (Zylo).
• 46% of licences go unused in the average company and that waste costs about $19.8M a year (Zylo).
• 25 to 60 tools make up most B2B marketing stacks with 35 to 45 being the common middle (Pedowitz Group).
• 34% of teams fully use their CRM and most use less than half its features (Insightly with Ascend2).
The CRM number doesn't tell the true story though. 55% of executives say their CRM is used effectively but only 27% of non-executives agree (MediaPost). It's the same story I hear from marketing leaders. Leadership approves the tools on a business case and then the people using them stop believing the data.
Which part of the stack is actually broken?
Martech covers a lot more than marketing automation. There's the CDP, analytics, web and sales enablement and most of those layers still do their job. The rest of this zooms in on one of them: the MAP and how it gets you from a buying group to revenue.
And interest in the MAP is falling fast. Branded search for the big four is down 24.9% over the last 12 months and 35.6% for the last three months against the same months last year (Storybook Marketing). Some of that drop is people asking AI tools instead of Google but it's still steep. Marketo, HubSpot, Pardot and Eloqua still hold 95% of the interest and the AI-native newcomers are only just showing up.
The reason goes back a long way. Sales used to own the relationship and brand's job was to open the door. Then buyers started researching online and left sellers out of the picture for a while. Sales needed help so marketing promised them one hand-raiser, the MQL, and we built a whole set of metrics around it. The software was built on that MQL thinking too. Lead-centric tools and lead-centric data models.
Then the playbook broke. Buying groups replaced the single lead and the handover to sales moved from MQLs to early opportunities. Account-based ways of working caught up with what sales had always done. Most MAPs didn't.
My thoughts: why is unused martech so hard to remove?
1. Tools arrive without an adoption plan. Gartner's respondents named ecosystem complexity and customer data problems as the biggest blockers to using more of their stack. Multi-year contracts then keep the unused tools on the books (Gartner).
2. Procurement protects the incumbent. A software contract takes 72 days on average from request to signature. That's 40 days for a new purchase and 82 for a renewal (Vertice). So buying the incumbent's add-on is quicker than buying something better. Jon Miller (Marketo's co-founder) calls the fear of switching that comes out of this "marketing automation Stockholm syndrome": teams stay loyal to a platform that holds them back because leaving feels riskier than staying (CMSWire).
3. The big platforms have every reason to keep you. They protect their subscription revenue with long contracts and a steady stream of AI add-ons, and a lot of those add-ons aren't ready yet. It makes it harder to tell what's proven and what isn't.
But here's a fourth reason and I believe it's the biggest one. Nothing in a modern stack stands on its own. The MAP syncs with the CRM and the CRM feeds audiences to the ad platforms. The webinar tool pushes registrations into both and an enrichment tool writes into the middle of all of it. It's a web of APIs. And rightly so because that's how data gets to where it's needed.
But it means no SaaS tool gets looked at on its own. Each one is a piece of a puzzle and if you take a piece away you lose part of the picture. A sync breaks or a lead-routing rule stops firing. A dashboard goes blank or a report finance relies on comes out wrong. So the unused tool stays because nobody's sure what it's still holding up. We see this in any long-running MAP instance where integrations get added every year and hardly ever removed.
Forrester makes the same diagnosis from the planning side. In Rani Salehi's ten non-negotiables for go-to-market planning the failure mode is disconnected systems held together by heavy customisation, so nobody gets a clear view of results. The enabler mode is one shared environment with governed automation.
“Technology is an enabler, not a solution.”
Rani Salehi, Principal Analyst, Forrester
Why do rules built for one lead break with a buying group?
Most MAPs were built on a simple if-then. If a lead visits the pricing page five times, score +20 and alert sales. That works when you're chasing one person.
A buying group doesn't work like that. If the CFO reads the business case and IT hasn't looked at security yet, send IT the security pack. But if IT has and procurement hasn't, do something else. And so on for every person in the group, depending on how far each of them has moved. Every extra person multiplies the combinations so you can't write all the if-thens. You need real-time decisions that read every signal and pick the next best action.
5 steps to a solution: how do you fix an underused or unused martech stack?
1. Audit usage before every renewal. Pull log-in and feature data four months out because renewals take the longest to get through procurement. Zylo suggests only keeping apps that run at 90% utilisation or more (Zylo). I'd start here rather than with a big consolidation project. An audit is a much easier conversation with a team that's already sceptical about tools. It's also the first stage of our go-to-market reviews.
2. Learn how your buyers buy. That's the outside view. Map who's in the buying group and how they move from first research through to renewal, because a MAP does more than win net-new clients. Agree which roles make up a typical buying group and what counts as a qualified one. Forrester's Conrad Mills calls this becoming "buying group capable". His point is that you start from the signals you already collect and package them into a group before anything goes to sales. Our ABM approach and guide to ABM cover how.
“Buying practices have changed faster than marketing and selling practices have.”
Conrad Mills, Principal Analyst, Forrester
3. Rebuild scoring, routing and decisioning. That's the inside view: how you manage all those interactions once they come in. Swap rigid rules for real-time decisions about who gets what next. Phave was built by Jon Miller and Nick Bonfiglio (who ran product at Marketo) and has been generally available since August 2026. It treats buying groups as records and uses AI to pick each person's next touch (Optimixed via MarTech). Bring it in for a problem you can name. Then run it on one segment next to your current platform before you migrate anything.
4. Start with your own data then add the rest. Get your first-party data right first: what buyers do with you on your website, in your emails, at your events, in your product and in your CRM history. Then enrich it with second-party data a partner shares with you and third-party data, like contact and company data from waterfall tools and account intent from providers like Bombora. Clay checks 150+ data providers in turn and writes the results into your CRM. Intercom moved its enrichment from several tools into Clay and grew outbound-sourced pipeline by 140% (Clay case study). It's also the quickest way I know back to CRM data people trust.
5. One in, two out. No new contract gets signed off unless it retires two existing ones by name.
If you want to start this week, list the martech renewals due in the next six months. Next to each one write down the last date it touched pipeline.
Clay vs Phave vs your current MAP: what does each layer do?
|
Tool |
Layer |
What it does |
What it can replace |
|
Clay |
Data and signals |
Waterfalls 150+ data providers and writes the results to the CRM |
Single-source enrichment contracts and manual list building |
|
Phave |
Decisions |
Treats buying groups as records and picks each person's next touch |
Consolidates your rules-based MAP, after a side-by-side pilot |
|
Your current MAP |
Rules-based execution |
Runs if-then campaigns built around individual contacts |
Stays in place until a pilot proves the switch |
|
CRM |
System of record |
Holds accounts, contacts, opportunities and activity history |
Nothing. It stays |
Phave's plans start at $36,000 a year and are priced on how many people get something from you each month (Phave). Clay prices its plans on two things: Actions for the platform work it runs, and Data Credits for data bought from its 150+ data partners. Plans start at a free tier. If you bring your own API keys for data you already license, you skip Data Credits and only use Actions (Clay).
How do Clay and Phave fit with the stack you already have?
They sit in different layers on top of the CRM and channels you already pay for. Clay answers who and when. Which accounts fit and who sits in the buying group. Whose details are out of date and what changed this week. Phave answers what next. Which person gets which touch and in what order, without two campaigns landing on the same CFO in one week.
Phave holds the first-party record for every person, account and buying group. A data lake attached to it brings in the third-party data from tools like Clay and Apollo so decisions run on all of it.
Each one does its own job well. Clay gives you well-researched lists and Phave makes good decisions on the data it's given. Put them together and the signal lands in the CRM and the next touch follows straight from it.
What does this look like in practice?
Here's a composite example. It's built from the kind of program we set up rather than one named client. A company sells production-planning software to 300 target manufacturers. Marketing and sales have agreed on a five-role buying group: operations, IT, finance, a plant manager and procurement.
Monday. Clay spots a new VP of Operations at one account plus four open production-planner roles. The CRM only has two of the five roles. Clay finds the other three and writes them into Salesforce with the signals that triggered it, so the account executive can see why.
Tuesday. Phave reads the account as a buying group with gaps. The new VP gets a short case study from a similar plant. The finance director is already in another campaign so Phave holds off for a week.
Thursday. Three of the five roles have engaged. Phave marks the group as qualified and the account executive gets the whole committee: who engaged and which role is still missing. The old set-up would have sent one MQL from whoever filled in a form.
Nothing that week needed a new channel. What left the stack was the standalone enrichment subscription and the spreadsheet someone was using to track buying-group coverage by hand.
Aren't Clay and Phave just two more apps on the pile?
Only if nothing leaves. Clay is there to retire single-source data contracts. Phave is built and priced to replace the marketing automation platform so you should end up with fewer tools than you started with. I'd still lead with the audit though. A new platform on top of a stack nobody trusts just makes the trust problem worse.
The bigger change is the unit of measurement and the timing makes it urgent. Forrester's Buyers' Journey Survey 2025 found 68% of B2B buyers already have a front-runner vendor in mind when the buying process starts. That front-runner wins 80% of the time (Forrester). An MQL from a form fill usually turns up after that choice has been made.
Most ABM programs still score individuals and hand sales an MQL. If the buying group is the unit, Clay fills in the committee and Phave qualifies the group. We've written about that shift as buying group marketing and about why signals need buyer research behind them.
Can Clay and Phave work without a CRM?
Possibly. I haven't seen it run this way in production yet but I can see it working for a small team starting from scratch. Say a scale-up has one or two sellers and no CRM. Clay builds the target list and finds the buying group. Phave holds the people, accounts and buying groups as records and picks the next touch. Sales works from the alerts in their inbox or Slack.
What you give up is the pipeline. There's no opportunity record and no forecast sales can stand behind, so revenue reporting lives in a spreadsheet. In my view this set-up works but isn't ideal. Phave is designed to bring calls, meetings and tasks in through a CRM sync, and it can create tasks for sales itself. So I'd treat it as a starting point and add a CRM once deals start piling up.
Who are Clay and Phave for, and who aren't they for?
Before I get accused of sounding like a sales rep for two SaaS companies, let me explain where this stack gets its value. And who it's for and who it's not.
Why it works
The value is in the hand-off. Clay keeps the CRM complete and current so Phave makes its decisions on data people trust. Phave turns that data into the next touch for each person in the buying group so Clay's research gets put to work straight away. Each tool is good on its own. Together each one makes the other more useful.
Who it's for and who it's not
|
|
It's a good fit if |
It's not for you (yet) if |
|
Budget |
You can fund Phave's list price plus Clay usage, ideally by retiring tools you already pay for |
The only way to pay for it is new budget on top of the current stack |
|
Current stack |
Your CRM is in reasonable shape and your MAP has become the bottleneck |
Nobody owns CRM data quality yet or your MAP went live in the last year and hasn't had a fair run |
|
Industry |
Complex B2B sales such as manufacturing, technology, financial services or professional services |
B2C, e-commerce or transactional sales with one decision maker |
|
Application |
Account-based programs with a defined target list and buying groups of several roles |
High-volume inbound where speed per lead matters more than the committee |
|
Experience |
Someone in marketing ops can own the data model and the rules |
Nobody owns the CRM day to day and there's no partner to cover it |
|
Risk |
You're happy to run a side-by-side pilot on one segment for a quarter |
You need a full migration signed off before anyone sees results |
|
Knowledge |
Sales and marketing agree (or are ready to agree) on what a qualified buying group is |
Sales and marketing still argue about what an MQL is |
If most of your answers sit in the right-hand column, start with the audit in step 1 and a go-to-market review. The tools can wait. If they sit on the left, our ABM and buying group marketing pages show how we set the program up. The Clay and Phave pages cover each tool.
Where do I think marketing automation could go next?
This is my opinion rather than anything a vendor has told me. Everyone's connecting AI agents right now. One agent finds the account news and another writes the email. Most of them do one job and start from scratch every time so they don't know where a buyer is in the journey or what the rest of the group has already done.
The option I keep coming back to is a shared memory layer. Something that holds every person and account across the whole journey so each agent picks up where the last one left off. The MAP could become that layer, or the CRM could, or something new could sit between them. I don't think anyone has fully solved it yet.
If you're testing agents today, I'd ask one question before you add another: what does this agent know about the buyer that the last one learned? If the answer is nothing, you're adding a tool that starts from zero every time.
Frequently asked questions
Is the SaaSpocalypse real?
The sell-off was real. About $1 trillion left software stocks in early 2026. The idea that companies will stop buying software is weaker because most marketing teams already pay for about twice what they use.
What does Forrester say about martech and buying groups?
Forrester principal analyst Rani Salehi lists technology as an enabler among ten non-negotiables for GTM planning. Her failure mode is disconnected systems held together by heavy customisation. Conrad Mills argues that MQL-based revenue processes are going obsolete because buying practices have changed faster than marketing and selling practices.
Does Clay replace data providers like ZoomInfo?
It can replace overlapping contracts because it checks many providers in turn. The licences you keep, ZoomInfo included, can be plugged into Clay so you still use them in the same workflows.
Can Phave run alongside Marketo or Eloqua?
Yes. Phave can run on one segment or region while your current platform keeps going so you can compare results before you migrate anything.
Can Clay and Phave work without a CRM?
Possibly for a small team starting from scratch, though I don't think it's ideal. Clay builds the list and Phave holds the people, accounts and buying groups. You give up pipeline records and forecasting, and sales activity is best logged through a CRM. So plan to add one once deals start piling up.
Who are Clay and Phave a good fit for?
Complex B2B sales teams running account-based programs with a CRM in reasonable shape and a MAP that has become the bottleneck. If your sales are transactional or nobody owns the CRM, start with a stack audit instead.
What should be cut from a martech stack first?
Overlapping data tools. In one of our current conversations, a 400-person B2B team catalogued 58 tools and found three separate audience stores. It retired eight low-adoption products.
Full disclosure: LeadFabric is a Clay Studio Partner and a European launch partner for Phave. We've also been implementing Marketo and Eloqua since 2008, so most of our work starts with deciding what a new tool replaces before anyone signs.
References
• Forrester (2026), Rani Salehi, Principal Analyst, presentation at Forrester B2B Forum EMEA, London
• Forrester (2026), Conrad Mills, Principal Analyst, presentation at Forrester B2B Forum EMEA, London
• Forrester (2025), Building preference is the key to winning B2B buyers, Buyers' Journey Survey 2025 data
• SoftwareSeni (2026), the SaaS reckoning explained
• Gartner (2023), martech utilisation drops to 33%
• Gartner (2025), marketing technology topic page, 49% utilisation
• Zylo (2026), SaaS licence statistics from the 2026 SaaS Management Index
• The Pedowitz Group, average number of tools in a B2B martech stack
• Insightly and Ascend2 (2025), CRM research report
• Vertice (2026), procurement cycle time
• Clay, Intercom customer story
• CMSWire (2026), Jon Miller launches Phave
• Phave, pricing and FAQ
• Storybook Marketing (2026), MAP category brand tracking, branded search volume to September 2026 (martech.org/author/liam-moroney)
Further reading from LeadFabric
• What is Clay and what does it mean for the future of B2B?
• Know your buyer, know the moment: closing the loop on signal-based growth
