Sales Hub Pipeline Setup: From Pipedrive Refugees to Forecast Accuracy

We’ve migrated 20+ Pipedrive instances to HubSpot. The single worst move every time is the same: lift-and-shift. The team exports the pipeline as-is, imports the deals as-is, recreates the same stages with the same names, and 60 days later the forecast still doesn’t work — because the problem was never Pipedrive. The problem was a sales process that Pipedrive had quietly let them fudge.

HubSpot is less forgiving. That is a feature, not a bug, but it surfaces problems that the migration is supposed to fix and most migrations don’t. Here are the six things to redesign on the way over.

Why Pipedrive’s pipeline structure doesn’t translate cleanly

Pipedrive is a deal-centric tool. The deal is the primary object; everything else (contacts, companies, activities) hangs off it. Pipelines are visual, drag-and-drop, and forgiving — you can move a deal to “Negotiation” with no required fields, no probability, no close date adjustment, and Pipedrive will not complain.

HubSpot is a contact-centric tool with deals attached. Deal stages have probabilities baked in. The forecast pulls from those probabilities. Required-field rules can stop a deal from progressing. Workflow triggers fire on stage changes. The same drag of a deal card causes 4–6 things to happen in HubSpot that did not happen in Pipedrive.

If your sales team has been moving deals freely in Pipedrive for years, those moves were probably wrong half the time and nobody felt it. In HubSpot, those wrong moves now produce wrong forecasts. The migration is the moment to clean this up.

Stage definitions — the 6-stage default and when to deviate

Most Pipedrive instances we inherit have between 8 and 12 stages. Most should have 6.

The 6-stage default we use for B2B SaaS:

Qualified to Buy — sales has accepted the lead, has confirmed budget exists, and has confirmed there is a real buying process. ~20% probability.
Discovery / Demo — first real conversation has happened. ~30%.
Proposal Sent — written proposal or quote is in the buyer’s hands. ~50%.
Negotiation — terms, contract, legal back-and-forth. ~70%.
Closed Won — signed. 100%.
Closed Lost — dead. 0%.

Some companies need a seventh stage between Discovery and Proposal — a “Technical Validation” or “Security Review” stage for SaaS with long IT-evaluation cycles. Add it only if the average deal sits in that phase for 14+ days. Otherwise the stage will compress and distort your funnel.

Stages we delete on every Pipedrive migration: “Lead In” (that’s a contact, not a deal), “Contact Made” (that’s an activity, not a stage), “First Email Sent” (also an activity), “Won — pending invoice” (use a deal property, not a stage). These stages exist in Pipedrive because Pipedrive’s flexibility encouraged them. They produce noisy pipelines and meaningless forecasts.

Property mapping that actually preserves history

The property mapping is the unglamorous part. Most migrations get this wrong by being too literal — every Pipedrive custom field gets a HubSpot equivalent, even the ones nobody used in three years.

The right approach: audit before you map.

Pull every Pipedrive deal field. Mark each as keep, merge, or drop.
Drop any field with <30% population across the last 12 months of deals. Nobody uses it.
Merge fields that are duplicates (“Industry” and “Vertical” become one).
Rename fields to match HubSpot conventions (snake_case for internal name, Title Case for label).

The single most-missed mapping: deal source. In Pipedrive this often lives as a free-text custom field. In HubSpot it should be a dropdown tied to your contact-source taxonomy. Migrating it as free text is what creates “Linkedin,” “LinkedIn,” “Linked In,” and “linkedin” all showing up as separate values in your reporting six months later.

History preservation is mostly about the activity log. Use HubSpot’s import-with-timestamps feature so closed-won deals from 2024 don’t all show as having closed on migration day. This is one line in the import config and most agencies skip it.

The deal-to-revenue chain that Pipedrive lets you fudge

In Pipedrive, the relationship between a deal amount, the products on the deal, the line items, and the actual contracted MRR/ARR is loose. You can have a deal at $50,000 with no products attached and Pipedrive doesn’t care.

HubSpot’s reporting, especially in Sales Hub Pro+, expects the chain to be tight: deal amount = sum of line items = recognized revenue (or its forecast). When the chain is broken, the forecast is broken, and your CRO will not trust the dashboard.

Three things to fix on the way over:

Line items, not deal amount. Build the price book. Migrate deals with line items attached, not flat dollar amounts. The first three weeks of this feel like overhead. Months 2 onwards, the forecast is dramatically more accurate.
MRR vs. ARR vs. one-time. Use HubSpot’s recurring revenue properties. Don’t smash all your revenue into a single number. The CFO will ask for the breakdown later and you don’t want to rebuild then.
Contract length and renewal date. Capture both at deal-creation time, not deal-close time. If you wait until close, your renewal pipeline three years from now is a manual reconstruction project.

Forecasting setup: probability vs. forecast amount vs. weighted pipeline

Pipedrive’s forecasting is mostly visual. HubSpot’s is mathematical. The three concepts you have to get straight on day one of the migration:

Probability. Set per stage, applies to every deal in that stage. This is the number HubSpot uses to compute weighted pipeline by default. Get it right per stage and most reps don’t have to think about it.

Forecast amount. A separate field a rep can set to override the deal amount × stage probability when they have specific intel. Useful for “I know this deal is going to close at $80K not the $100K on the contract” or “I know this is dead even though it’s still in Proposal stage.” Most reps will not use this; that’s fine. Make it available, train on it, but don’t require it.

Weighted pipeline. The math: deal amount × stage probability, summed. This is what your sales leader looks at on Monday morning. It should match the rep-level forecast within 10–15%. If it doesn’t, either probabilities are wrong or reps are gaming forecast amounts.

The mistake we see most often: companies migrating from Pipedrive copy stage probabilities directly. Pipedrive defaults are “the way our last sales leader felt about each stage in 2022.” HubSpot defaults are also wrong. Re-derive your probabilities from your last 12 months of historical conversion data. Do this in week 1 of the migration, not week 6.

Sales rep change-management: the 2-week onboarding plan that sticks

The hardest part of any Pipedrive-to-HubSpot migration is not the data; it is the reps. Pipedrive has lower friction. Reps liked it. They will not love HubSpot in week one and they will tell you so.

The 2-week onboarding plan that actually works:

Week 1, days 1–3: in-person or live-video training, 90 minutes per day. Not pre-recorded. Not “here’s a Loom and your login.” Live, with the implementer, with reps’ actual deals on screen, in their actual portal.

Week 1, days 4–5: shadowing. The implementer sits in (virtually) on a sales call per rep, then watches them update HubSpot afterwards. Catches the muscle-memory mistakes (forgetting to log the call, skipping required fields) before they become habits.

Week 2, days 1–5: daily 15-minute office hours. The implementer is on a Zoom or Slack channel for 15 minutes a day. Reps drop in with questions. Most days nobody shows up. The days reps do show up are the days you save 6 weeks of bad habits.

End of week 2: the dashboard sign-off. Each rep walks the sales leader through their pipeline in HubSpot, on screen. The reps who haven’t adopted will be visibly uncomfortable. Address it then, not at month 3.

This is 4–6 hours of senior implementer time per rep across two weeks. It is the single highest-ROI investment in the entire migration. Skip it and your $40K migration produces a portal nobody uses. Do it and your CRM adoption stays above 90% in month 4 — the threshold above which forecasts become trustworthy.

What to do next

If you are scoping a Pipedrive migration: don’t sign anything until your prospective partner has answered “what would you delete from our current pipeline structure” with specifics. A partner who doesn’t push back on stage count is a partner who will lift-and-shift you.

If you are 30 days into a migration that feels off: audit the stages and the deal-source field first. Those two things break 80% of post-migration forecasts. Our HubSpot audit checklist covers the full framework, and the workflow performance audit covers the automation side.

If you would rather we ran the migration: book a free 30-min consultation. We’ll send you the 6-stage redesign for your specific deal flow before you sign — a one-hour scoping call usually surfaces the three things lift-and-shift would have broken.