Your marketing team wants multi-touch attribution. Your sales team wants leads routed instantly. Someone has found a vendor for each problem. LeanData for routing. Bizible or HockeyStack for attribution. Maybe CaliberMind or Dreamdata on top.

The combined annual cost? Somewhere between £40k and £80k.

For lead routing and attribution. Two problems that can be solved architecturally inside your existing CRM for a fraction of that spend.

I'm not saying these tools don't work. They do. I'm saying most companies buying them haven't exhausted what their existing systems can do — and the tool purchase is masking a governance problem that the tool won't fix.

The Lead Routing Problem Is Simpler Than Vendors Claim

Lead routing vendors sell complexity. Round-robin algorithms. Territory-based matching. Account-based routing with enrichment triggers. Weighted distribution with capacity balancing.

Sounds sophisticated. Now ask yourself: how many of those features do you actually need?

Most B2B SaaS companies between £2M and £30M ARR have a routing model that boils down to three rules:

  • If the lead matches an existing account, route to the account owner.
  • If the lead is in a named territory, route to the territory owner.
  • If neither applies, round-robin across the available pool.

That's it. Three rules. You don't need a £20k/year platform to enforce three rules.

What you actually need

A lead routing system that works requires four things. None of them are a dedicated routing tool.

  • A clean account-to-territory mapping. Every account assigned to an owner. Every territory defined. This is a CRM data quality exercise, not a software purchase.
  • Lead-to-account matching logic. When a new lead arrives, match it against existing accounts by email domain, company name, or enrichment data. Salesforce can do this natively with matching rules. HubSpot can do it with workflows and company associations.
  • Assignment rules or flows. The actual routing logic. In Salesforce, this is Lead Assignment Rules or Flow Builder. In HubSpot, it's workflow-based assignment. Both handle round-robin, territory matching, and account-based routing.
  • Speed-to-lead tracking. How quickly is a routed lead being contacted? This is a timestamp comparison — lead created timestamp versus first activity timestamp. A formula field and a report. Not a platform.

The routing "problem" that vendors sell is 80% data quality and 20% automation logic. If your account data is clean and your territories are defined, the routing practically builds itself.

Where Routing Actually Breaks

When lead routing fails, companies blame the routing logic. It's almost never the logic. It's one of three things.

Dirty account data

Duplicate accounts. Accounts without owners. Territories that overlap. When the foundational data is messy, no routing tool can produce clean outcomes. It'll just route faster to the wrong person.

Fix the data. The routing follows.

Undefined territory rules

Who owns what? If the answer requires checking a spreadsheet, asking a manager, or "it depends on the relationship," you don't have a territory model. You have a negotiation.

Territory rules need to be explicit, documented, and encoded in the CRM. Geographic, segment-based, named account — pick a model and enforce it. The routing is just the automated expression of the territory logic.

No accountability for follow-up

The lead gets routed. Nobody contacts it for 48 hours. The lead goes cold. Marketing blames sales. Sales blames lead quality.

The fix isn't faster routing. It's follow-up governance.

  • SLA on first contact: 4 hours for inbound demo requests, 24 hours for content leads.
  • Automated escalation: if the SLA is missed, the lead re-routes to a backup or triggers a manager alert.
  • Tracking: speed-to-lead metrics visible to marketing and sales leadership. Shared accountability.

Lead routing is a governance problem with a thin automation layer on top. Not the other way around.

The Attribution Trap

Now let's talk about the more expensive problem. Attribution.

Marketing attribution has become a cottage industry. Multi-touch attribution platforms promise to tell you exactly which touchpoints influenced each deal. First touch. Last touch. Linear. Time-decay. W-shaped. U-shaped. Full-path.

The models multiply. The dashboards proliferate. And the CMO still can't confidently answer the question: "Which channels should we invest more in?"

Here's why.

Perfect attribution is a fantasy

A B2B buying journey involves 6–10 stakeholders, multiple channels, offline conversations, peer recommendations, analyst reports, and Google searches you'll never track. The idea that a software model can assign fractional credit to each touchpoint with meaningful accuracy is a comforting fiction.

Multi-touch attribution models don't reveal truth. They distribute credit according to a formula that the model designer chose. Change the model, change the results. Same data, different story.

A CMO using W-shaped attribution will make different budget decisions than one using time-decay — not because the underlying reality is different, but because the model weights touchpoints differently.

That's not insight. That's an assumption disguised as analysis.

The data is always incomplete

Attribution models require tracking every interaction across every channel. In practice:

  • Dark social is invisible. That LinkedIn post someone shared in a Slack channel? Untrackable.
  • Offline conversations don't register. The CRO mentioned your product at a dinner. The prospect Googled you the next day. Your model credits Google.
  • Multi-stakeholder buying means different people interact with different channels. The champion read your blog. The CFO saw the case study. The VP attended a webinar. Your attribution model sees three separate journeys, not one deal.
  • Cookie deprecation and privacy changes are making digital tracking less reliable every year, not more.

Attribution platforms process the data they can see. But the data they can't see — the dark funnel — often influences the deal more than the trackable touchpoints.

What to Build Instead of an Attribution Platform

You need attribution. You just don't need the version the vendors are selling.

Here's a pragmatic attribution architecture that costs nothing beyond CRM configuration time and answers the questions that actually matter.

First-touch source tracking

Capture one thing reliably: how did this lead first enter your system?

  • Organic search
  • Paid search
  • Paid social
  • Organic social
  • Referral
  • Event
  • Outbound (SDR-sourced)
  • Partner
  • Direct / unknown

This is a single field on the lead record, set at creation, never overwritten. Populate it from UTM parameters, form submissions, or manual entry for offline sources.

First-touch source carried through to the opportunity and closed-won deal gives you the most important attribution metric: which channels create pipeline that actually closes?

Not which channels generate the most leads. Which channels generate revenue. The distinction matters enormously for budget allocation.

Self-reported attribution

Add one open-text field to your demo request or contact form: "How did you hear about us?"

This captures what no tracking pixel ever will. The prospect tells you: "My colleague mentioned you." "I saw your LinkedIn post." "You were recommended in a Slack community."

Self-reported attribution won't be complete or perfectly accurate. But it captures dark funnel signals that digital attribution models miss entirely. And the data compounds over time — after six months, you'll see clear patterns about which organic and word-of-mouth channels are actually driving awareness.

The combination of first-touch digital tracking and self-reported attribution gives you 80% of the insight at 0% of the platform cost.

Campaign-to-opportunity tracking

Salesforce has native campaign influence tracking. HubSpot tracks campaign associations automatically. Use what's already there.

Attach campaigns to opportunities. Track which campaigns touched which deals. You don't need multi-touch fractional credit models. You need to know: "Of the deals that closed this quarter, which campaigns were involved?"

A simple campaign influence report tells the CMO which programmes are appearing in closed-won deals. That's more actionable than a fractional credit model that distributes 7.3% of a deal to a webinar attended four months ago.

Cohort-based channel analysis

Instead of attributing individual deals, analyse channels in cohorts.

  • Leads sourced from paid search in Q2: what was their conversion rate to opportunity? To closed-won? What was the average deal size?
  • Leads sourced from content marketing in Q2: same analysis.
  • Leads sourced from events in Q2: same.

Compare the cohorts. Which channels produce leads that convert at higher rates, close at higher values, and retain better?

This doesn't require a platform. It requires clean first-touch data and a reporting layer that can slice by source. Every CRM does this natively.

Cohort analysis answers the budget question better than multi-touch attribution because it measures outcomes, not touchpoints.

The Architecture That Replaces Both Tools

Here's the full build. Lead routing and attribution, CRM-native, no overlay platforms.

Data foundation

  • Clean account data with assigned owners and defined territories.
  • Standardised lead source taxonomy (the list above, or your version of it).
  • UTM parameter capture on all forms, flowing into a persistent first-touch source field.
  • Self-reported attribution field on high-intent forms.

Routing layer

  • Lead-to-account matching (native matching rules or a lightweight enrichment step).
  • Territory-based assignment rules.
  • Round-robin fallback for unmatched leads.
  • SLA automation: escalation triggers if first contact doesn't happen within the defined window.

Attribution layer

  • First-touch source persisted from lead to opportunity to closed-won.
  • Self-reported attribution captured and categorised.
  • Campaign influence tracking using native CRM capabilities.
  • Quarterly cohort analysis by source: conversion rates, deal size, sales cycle, and retention.

Reporting layer

  • Pipeline generation by source (first-touch) — weekly.
  • Speed-to-lead by routing path — weekly.
  • Campaign influence on closed-won revenue — monthly.
  • Cohort performance by channel — quarterly.
  • Self-reported attribution trends — quarterly.

Total build time: two to three weeks of RevOps effort. Total ongoing cost: zero platform fees. Just the CRM you're already paying for.

When the Overlay Is Justified

Fair is fair. There are scenarios where dedicated tools earn their keep.

  • Routing at extreme volume. If you're processing thousands of inbound leads per day with complex account hierarchies and real-time enrichment requirements, a dedicated routing engine like LeanData handles edge cases that CRM-native rules can't. Most companies aren't at this volume. If you're processing 50 leads a day, native rules are fine.
  • Enterprise-scale attribution with board-level reporting requirements. If your board demands granular multi-touch models and your marketing budget exceeds £2M annually, the investment in a dedicated platform may be warranted — but only after first-touch tracking and self-reported attribution are already in place.
  • Complex ABM routing. If you're running a mature account-based marketing programme with intent data triggers, multi-threading across buying committees, and real-time enrichment-based routing, the native CRM capabilities may fall short.

The pattern is the same as every other tool in the revenue stack: build the governance first, exhaust the native capabilities, and buy the overlay only when you've genuinely outgrown what you have.

The Budget Reallocation

Let's do the maths.

A typical mid-market company spending on routing and attribution tools:

  • Lead routing platform: £15k–£25k/yr
  • Attribution platform: £25k–£50k/yr
  • Implementation and integration: £10k–£20k (often recurring as the stack changes)
  • Admin overhead: 15–25% of a RevOps person's time maintaining both systems

Total: £50k–£95k annually.

The CRM-native alternative costs two to three weeks of RevOps build time upfront and minimal ongoing maintenance.

The delta? Invest it in the thing that actually drives pipeline: better content, more targeted campaigns, or an additional SDR. The money you save on routing and attribution platforms can fund the marketing programmes that those platforms were supposed to measure.

There's a certain irony in spending £50k to measure the effectiveness of marketing programmes that would be more effective if you'd spent that £50k on the programmes themselves.

Build the architecture. Save the budget. Invest in what actually generates revenue.