A Masters degree student pursuing a data science degree from Notre Dame recently asked me five questions about my knowledge of forecast measurement and the B2B SaaS industry.  The answers take my experiences built over the years into account, they are not AI generated (but we used AI to proofread for errors).

 

1. How do clients measure whether their marketing and sales funnel is healthy?

 

Most clients use a combination of platform dashboards (Demandbase, 6sense, Salesforce) to track account engagement, velocity through funnel stages, and conversion rates (we advocate for one dashboard in Salesforce). Key metrics include:

 

•   Account engagement (intent signals, website visits, campaign responses)
•   Stage velocity (how quickly accounts move from engagement to opportunity to close)
•   Marketing-sourced and marketing-influenced pipeline
•   Custom dashboards in Salesforce or Power BI for unified reporting
•   Baseline cohort analysis: comparing accounts with engagement vs. those without, to show impact on meetings and pipeline creation.

 

2. Where are the biggest gaps when integrating data from 6sense, Demandbase, Marketo, Salesforce, etc.?

 

• Data hygiene and integrity: Inconsistent or incomplete data in Salesforce or Marketo undermines unified views.
• Lead-to-account matching: Difficulty connecting leads from marketing automation to the right accounts in Salesforce. (3rd party solutions often enter the picture here)
• Custom object limitations: Most platforms can’t update custom objects, requiring manual workarounds. (some 3rd party integrations do work here but fewer than expected).
• API and integration complexity: Each tool has its own data model, and syncing custom fields or historical data can be challenging.
• Real-time data lag: Some platforms update in near real-time, others batch sync, causing reporting delays.
• Siloed intent and engagement data: Not all platforms ingest third-party or anonymous data equally, leading to blind spots.

 

3. What is the most common reason a pipeline forecast ends up being off?

 

• Missing or misinterpreted intent signals: Not all buying signals are captured, especially from anonymous or “dark funnel” activity.
• Data lag: Delays in syncing data between systems (e.g., Marketo to Salesforce) can cause outdated forecasts.
• Poor sales activity logging: If reps don’t log activities, engagement and pipeline health are misrepresented. This is where being ruthless as a leader would help clean up bad forecasts.
• Overreliance on engagement scores: Predictive models may overvalue engagement without considering fit or historical close rates.
• Model drift: Predictive models become less accurate if not retrained as market conditions or buyer behavior changes.

 

4. How often do clients revisit or retrain predictive models/account scoring logic, and what triggers it?

 

• Most clients revisit or retrain models quarterly, or when there’s a major change in go-to-market strategy, product, or target market.
• Triggers include: significant shifts in win rates, new product launches, entering new markets, or observed model drift (e.g., declining forecast accuracy).
• Some platforms (like 6sense) offer quarterly model refreshes and benchmark reports to track model performance over time.

 

5. If you had a magic wand, what one data or insight capability would you immediately add to your stack?

 

• Real-time, unified account view: Seamless integration of all engagement, intent, and sales activity data (including anonymous signals) into a single, actionable dashboard.
• Granular attribution: Ability to trace every touchpoint and its influence on pipeline and revenue, across all platforms.
• Predictive “next best action” recommendations: AI-driven, context-aware suggestions for sales and marketing, based on the latest signals and historical outcomes
• Cheap option – the current platforms are waaaay too expensive for this economic environment and the appetite of non-sales and marketing leadership (eg boards, CFO, and CEO).

 

So to summarize, accurate pipeline forecasting in B2B SaaS requires a dual focus on both data integrity and disciplined sales management. On the technical side, common pitfalls include missing or misinterpreted intent signals, data lag between systems, and overreliance on engagement scores or outdated predictive models. Equally important are human factors: over-optimism, lack of intellectual honesty in deal stages, and poor pipeline hygiene—such as allowing stale or misclassified deals to persist. The most reliable forecasts come from organizations that combine clean, integrated data with rigorous, honest deal inspection and active pipeline management. Addressing both dimensions is essential for forecasts that truly reflect reality and drive better business decisions.