--- title: "Pipeline by Source: A Five-Day Plan to See Where Opportunities Really Start" description: "When marketing, SDRs and AEs each claim the same deals and the CRM says Other, hours get allocated by argument. This five-day plan builds a source report you can defend and a rule for moving hours." author: "Jamie Mitchell" category: "Pipeline Generation" date: 2026-11-05T06:50:05.726Z canonical: "https://salesbrew.co/blog/pipeline-by-source-how-to-track-where-opportunities-come-from" --- # Pipeline by Source: A Five-Day Plan to See Where Opportunities Really Start ![a train traveling through a forest filled with lots of trees](https://images.unsplash.com/photo-1714901423336-1884cd3fb50f?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=M3w4OTQwNjJ8MHwxfHNlYXJjaHwxfHxQaXBlbGluZSUyMGJ5JTIwU291cmNlJTIwYnVpbGQlMjBzb3VyY2UlMjByZXBvcnR8ZW58MXwwfHx8MTc5MTgwNDAyNnww&ixlib=rb-4.1.0&q=75&w=1200&auto=format) > When marketing, SDRs and AEs each claim the same deals and the CRM says Other, hours get allocated by argument. This five-day plan builds a source report you can defend and a rule for moving hours. Put the question to five people and you collect five answers. Marketing credits campaigns, SDRs credit sequences, AEs credit their networks, and the CRM shrugs with "Other". Until data settles it, every decision about where to spend hours is really a negotiation over credit. In five sessions of about an hour you can build a source report that holds up: a controlled source field, two quarters of cleaned history and a yield figure for each source. The end product is one decision, which source gets more of the team's hours and which gets fewer. ## The Week Mapped Out SessionOutputTest it passes Day 1A list of sources and one credit ruleTwo people tag the same five deals identically Day 2Two quarters of history, taggedFewer than a fifth of deals left unclear Day 3Yield per sourceYou can name the best and worst source Day 4A locked field and a shared-credit ruleReps cannot edit the source after creation Day 5A review and a decision about hoursOne source to grow, one to trim It is a stricter cousin of the weekly time split in [where pipeline comes from](/pipeline-generation-where-pipeline-comes-from-and-how-to-split-your-week). ## Day 1: Write the Dictionary List six to eight sources. One sensible opening set: demo requests, content and search, your own events, SDR outbound, AE-sourced, partners, customer referrals and expansion. Keep Other purely as an error bucket that you plan to drain. Salesloft's 2024 write-up of an Insight survey advises teams to [track pipeline contribution by source and then go more granular](https://www.salesloft.com/resources/blog/maximize-sales-pipeline-and-forecast-accurately), for example splitting demo requests from owned events and paid search. Start coarse. A source earns a split only after it accumulates enough deals to make two lines meaningful. Then fix the credit rule in one sentence. An example: credit goes to the earliest interaction that produced a genuine conversation with the buyer, and the final touch before record creation gets none. Both first-touch and last-touch views distort, so the goal is uniform use, not perfection. ## Day 2: Rebuild the Last Six Months Extract all opportunities created over the past half-year, with owner, creation date, existing source, stage and amount. For blanks, read the account timeline: earliest meeting invite, earliest form fill, earliest intro email. Count the records you cannot place. Should over a fifth remain unclear, tell leadership the output is directional and address the process on Day 4. Remember that a backfill documents history; it cannot tell you what would have happened without a given source. ## Day 3: Work Out Yield Per source, calculate opportunities created, win rate, mean deal size, median cycle and hours invested. Yield is opportunities, or won revenue, over hours. Precise hours are not needed; ask each owner for a candid weekly estimate. The first four are the inputs of [HubSpot's sales velocity formula](https://blog.hubspot.com/sales/sales-velocity), and the [velocity guide](/sales-pipeline-velocity-formula-and-the-lever-to-pull-first) shows the arithmetic. Salesloft's editors say that inside their own company the fastest, highest-converting pipeline is the AE-sourced kind. That describes a single firm, and computing your own figure is how you find out if you match. The sums also stay silent on saturation, because a source that shines at ten deals a quarter could stall at twenty. ## Day 4: Lock the Field - Require source when an opportunity is created. - Lock it afterwards, so only RevOps can amend it. - Decide how shared work is credited. The Salesloft article describes a lift when AEs and SDRs [prospect into the same accounts in a coordinated manner](https://www.salesloft.com/resources/blog/maximize-sales-pipeline-and-forecast-accurately). Give the deal one source for credit and use a second field for whoever assisted. - Add a field naming the partner or referrer, so the sender is never lost. Reps may still game the field. Audit ten deals a month. ## Day 5: Review and Move Hours Bring the table to sales, marketing and the SDR lead with three prompts: what returns the most per hour, what returns the least, and how would we spend a tenth of the lowest performer's time? Move that time to the leader for a quarter and re-measure. A single quarter with eight deals is too thin to justify abandoning a source. [The target account list plan](/how-to-build-a-target-account-list-in-five-days) shows where to aim the extra time. ## Ask the Buyer as Well as the CRM A CRM source field records what your systems saw. It cannot see the podcast the buyer heard, the peer who recommended you or the Slack thread that put you on a shortlist. Those influences shape preference long before a form is filled, and 6sense's 2025 report says [buyers chose from their Day One shortlist 95% of the time](https://6sense.com/science-of-b2b/buyer-experience-report-2025/), which means much of the decision precedes anything you can log. So add one open-text question to the first call or demo form: how did you first hear about us? Store it in a separate field and compare it with the system source each quarter. Where the two disagree, believe neither blindly, but treat the gap as a measure of how much of your demand is invisible. I would not build multi-touch attribution on top of this; a clear rule and a second, honest field are easier to audit. The idea fails for a tiny team: thirty opportunities of free-text answers is mostly noise. Collect them anyway and hold your conclusions until volume builds. Our [sourced and influenced rules](/sourced-vs-influenced-pipeline-definitions-and-the-rules-that-keep-them-honest) help settle disputes over credit. Share the first version of the report with a short note on how it was built: the sources, the credit rule, the date range and the number of deals left unclassified. A report that arrives with its method attached gets challenged on its logic and not on its authenticity, which is the argument you want to have. Keep the note in the same file as the table so it travels with it. ## Moves That Waste the Week - **Multi-touch models.** They appear rigorous but are hard to check. A single plain rule earns more trust. - **Twenty sources.** Tiny buckets produce noise, not insight. - **Leaderboards on source.** You want honest tagging, and a ranking rewards the opposite. ## Places It Breaks - **Slow cycles.** At nine months, win rate per source requires a long window. Rely on creation counts and stage conversion in the meantime. - **Brand halo.** Content and events build preference before any form is filled, so a deal logged as direct may owe a debt to content. Treat those channels' numbers as minimums. - **Few deals.** Below thirty opportunities in the window, any trend is fragile. ## Questions on source reporting ### What does a pipeline by source report show? A split of opportunities by their origin, whether inbound, SDR outbound, AE-sourced, partner or referral. ### How many sources should be tracked? Start with six to eight, and divide a source only when its volume can bear it. ### First touch or last touch? Either, provided you stay consistent. The first meaningful interaction is a good default, paired with an assisting-role field. ### How often should the report be reviewed? Check data quality every month and revisit hour allocation every quarter. ## The Test of Success Success looks like two leaders citing one source mix in a single meeting. Beyond that, treat the report as a tool for deciding, not a scoreboard. The full system is in the [pipeline generation guide](/b2b-pipeline-generation-complete-guide). ## FAQ ### What is a pipeline by source report? It breaks opportunities down by where they began, such as inbound, SDR outbound, AE self-sourced, partner or referral. ### How many sources should be tracked? Six to eight to begin. Split one only when it holds enough deals to merit a line. ### Should credit follow the first touch or the last? Pick one and keep it. The first meaningful interaction works well, with an assisting-role field for shared credit. ### How often should the report be reviewed? Monthly for data quality and quarterly for how hours are allocated. --- Source: https://salesbrew.co/blog/pipeline-by-source-how-to-track-where-opportunities-come-from