B2B pipeline planning: connect your revenue target to delivery economics

How to use this guide

A revenue target is not an operating plan until you can explain the customers, opportunities, conversations, time and delivery capacity behind it. For a company entering Japan or another APAC market, the arithmetic is especially useful because early assumptions are often imported from a more established market.

This guide provides a planning model, not an industry benchmark or financial forecast. All amounts and conversion rates below are hypothetical. Replace them with consistently defined cohort data when it becomes available, and keep new-market assumptions separate from established-market results.

1. Define what the target actually measures

Start with one unit: new contracted annual recurring revenue (ARR), bookings, recognized revenue or cash collected. These are different quantities. The example here uses new contracted ARR from recurring subscriptions, excluding one-time implementation fees. It is a planning definition, not an accounting policy. Have your finance team determine revenue recognition and reporting treatment.

Suppose the goal is ¥24 million in new ARR and the average recurring annual contract is ¥3 million. The model needs eight new customers: ¥24 million divided by ¥3 million. If contracts begin throughout the year, ¥24 million of new ARR does not mean ¥24 million of revenue recognized during that year. If customers pay later, it does not mean that much cash has arrived either.

Define an opportunity before calculating pipeline. In this model, a qualified opportunity requires a confirmed use case, an accountable buyer contact, a credible evaluation or purchasing path, a next action accepted by both sides and an amount basis. A downloaded PDF or a pleasant first meeting is not enough.

2. Work backward through explicit stage definitions

Assume, for illustration, that 25% of qualified opportunities become customers, 40% of completed discovery meetings become qualified opportunities and 10% of contacted target accounts complete a discovery meeting within the observation window.

The implied requirements are:

• Eight wins divided by 25% = 32 qualified opportunities.
• 32 opportunities divided by 40% = 80 completed discovery meetings.
• 80 meetings divided by 10% = 800 contacted target accounts.

At ¥3 million per opportunity, 32 opportunities represent ¥96 million of unweighted potential ARR. That is four times the target in this specific example because the assumed win rate is 25%. “Four times pipeline” is not a universal requirement.

Use unique accounts for the account denominator, not emails sent or contacts added. If several people at one company attend the same meeting, do not count them as several opportunities. Define how separate buying units or distinct opportunities within one company are handled. Keep inbound, outbound and partner-originated cohorts separate before combining them into a capacity plan.

This simple funnel assumes one primary opportunity per contacted account and a consistent offer. If your business has multiple products, expansion revenue or multiple opportunities per account, model those flows separately.

3. Put time back into the equation

A funnel without dates can produce a convincing annual total that cannot arrive before the deadline. If you assume six months from qualification to signature, opportunities created near year-end should not be counted as year-end wins without supporting evidence.

For each cohort, record first contact, completed discovery, qualification, evaluation start, decision date and actual close or loss. Separate open opportunities from completed ones when estimating conversion. A new cohort full of still-open deals cannot be compared fairly with a mature cohort whose outcomes are known.

Build the operating plan by month: starting opportunities, newly qualified opportunities, opportunities reaching a decision, wins, losses and ending open opportunities. Use an explicit “later than planned” outcome. Moving every delayed deal into next month preserves the spreadsheet total but hides the weakness of the forecast.

Where data is sparse, show a range instead of a precise weighted forecast. A stage probability assigned by a salesperson is an assumption until it has been tested against enough comparable outcomes. Do not treat the sum of weighted opportunities as guaranteed bookings.

4. Stress-test the uncertain inputs

Keep the eight-customer goal and ¥3 million contract value fixed, then compare three hypothetical conversion scenarios:

• Cautious: 15% opportunity-to-win, 30% discovery-to-opportunity, 5% account-to-discovery. Round up at each stage: 54 opportunities, 180 meetings and 3,600 contacted accounts.
• Working case: 25%, 40% and 10%. Requirements: 32 opportunities, 80 meetings and 800 accounts.
• Strong: 35%, 50% and 15%. Round up at each stage: 23 opportunities, 46 meetings and 307 accounts.

These are sensitivity tests, not forecasts or market benchmarks. Their purpose is to expose the decision that matters. If the addressable account list contains only 250 plausible companies, a plan requiring 3,600 unique accounts fails before considering sending volume. You may need a different segment, higher contract value, stronger access to buyers or a lower near-term target.

Change one assumption at a time as well. That shows whether improving buyer access, qualification or closing would have the greatest effect. Avoid “solving” the plan by raising every conversion rate simultaneously.

5. Check whether the team can deliver the funnel

Convert activity into workload. In the working case, suppose each of the 80 discovery meetings requires two total staff-hours including preparation and follow-up, and each of the 32 qualified opportunities needs a further 12 presales engineering hours. That is 160 plus 384, or 544 staff-hours before onboarding, general prospecting, administration or unsuccessful technical work outside that assumption.

Now add eight implementations at 40 hours each: 320 more hours. These illustrative inputs already total 864 hours. Assign the work by role and month, not only as an annual team total. A single specialist may become the bottleneck even if the organization appears to have spare aggregate capacity.

Include customer obligations too. A pilot that requires substantial integration work from the buyer may reduce participation or delay the schedule. Record evaluation burden when interviewing buyers, and reduce scope where possible rather than assuming local prospects will accept the same process as existing customers.

6. Measure acquisition cost and contribution consistently

For planning, define customer acquisition cost (CAC) as attributable sales and marketing acquisition costs divided by new customers acquired. Include a consistent allocation of salaries, tools, agencies, travel, partner acquisition fees and presales effort. State what is excluded and why. Match costs to the customers or acquisition period they support; dividing this month’s costs by this month’s wins can distort a long-cycle business.

If the relevant acquisition costs are ¥12 million for eight customers, illustrative CAC is ¥1.5 million per customer. Keep one-time market-entry research and setup visible as a separate investment line, then show a fully loaded view including it. Neither hiding setup costs nor mixing them into every future customer is a useful planning shortcut.

For a simplified subscription example, ¥3 million of annual recurring revenue equals ¥250,000 per month. At an assumed 70% recurring gross margin after the delivery costs included in your definition, monthly recurring gross profit is ¥175,000. Simplified CAC payback is ¥1.5 million divided by ¥175,000, or about 8.6 months.

This is not cash breakeven. It excludes churn, expansion, collection timing, financing, taxes, discounting and any costs omitted from the stated definitions. If onboarding costs are not already included, subtract onboarding contribution or add its net cost in a separate payback scenario. Avoid double-counting implementation labor in both CAC and delivery costs. Choose a consistent classification with finance.

For AI offers, test usage-cost exposure explicitly. A flat subscription price paired with variable inference or support costs can produce very different margins across customers. Model a normal-use and high-use scenario, including the commercial treatment of overages. For a partner route, model the actual invoicing and revenue arrangement, partner share, enablement and support burden without deducting the same fee twice.

7. Do not manufacture lifetime value from immature cohorts

A frequently used shortcut divides periodic customer gross profit by periodic churn. It relies on simplifying assumptions about stable churn and customer economics. A new market with a few customers and little renewal history cannot support that precision.

Instead, show contribution over explicit 12- and 24-month scenarios, with renewal, expansion and delivery assumptions visible. Distinguish contracted recurring value from revenue that depends on a future renewal. Include a downside case in which a customer does not renew and a large evaluation does not convert. This makes the amount at risk easier to understand than a confident lifetime-value ratio built on untested inputs.

8. Review one dashboard, make one decision

Review account fit, completed discovery, qualified opportunities, stage aging, wins, acquisition spending, presales hours and delivery contribution together. Use a consistent cohort and currency basis. Keep actuals, assumptions and targets visually distinct.

If meetings rise but qualification falls, tighten targeting or discovery. If qualified pipeline grows but decisions slip, investigate the evaluation and buying process. If wins improve while delivery contribution falls, revisit scope, price and service design before scaling acquisition.

The useful output is not a larger pipeline number. It is a decision about what to improve next, what the team can afford to test and which assumptions must be verified before increasing spend.