Seeds, Nets, and Spears: The Lead Source Mix That Makes a Pipeline Predictable

A predictable pipeline runs on three separate lead engines: Seeds (referrals), Nets (inbound), and Spears (outbound), a framework from Aaron Ross's Predictable Revenue. Referrals convert best, NAR data shows 40 percent of buyers find their agent that way, but only a deliberate 20/20/60 mix gives you control.

A pipeline becomes predictable when it runs on three deliberately separate lead engines: Seeds, the referrals and relationships that convert best; Nets, the inbound marketing that scales widest; and Spears, the targeted outbound you control most directly. The framework comes from Aaron Ross and Marylou Tyler's book Predictable Revenue (2011), and it remains the cleanest mental model for diagnosing why a real estate team's revenue swings from feast to famine.

Most luxury agents do not have a lead problem. They have a lead concentration problem. Their business runs almost entirely on one engine, usually Seeds, and the engine is invisible, untracked, and unmanaged. This article breaks down what the data says about each source, why they demand separate workflows, and how a 20/20/60 mix turns a talented closer into the owner of a controllable growth system.

What are Seeds, Nets, and Spears?

In Predictable Revenue, written from the playbook Aaron Ross built at Salesforce, where his outbound process added roughly $100 million in recurring revenue, per the book's own account (2011), leads are sorted by how they are created, not by who they are. Seeds are relationship-driven: referrals from past clients, word of mouth, professional networks, sphere of influence. Nets are marketing-driven: you cast a wide net through content, social media, portals, events, and paid campaigns, and you catch whoever swims in. Spears are effort-driven: a human deliberately targets a specific prospect, an expired luxury listing, a developer who needs a sales operation, a specific building's owner list, and hunts.

The categories matter because each engine has different economics, different timelines, and different failure modes. Seeds convert at the highest rate, arrive with trust pre-installed, and cost almost nothing, but they scale with your relationship history, which means slowly, and you cannot order more of them this month. Nets scale widely and build brand while you sleep, but they convert at low rates, take months to compound, and attract unqualified volume that can drown a team. Spears are the only engine you can point: you choose the target, the message, and the volume of attempts, which makes outbound the only source that responds proportionally to effort this week. Ross's core argument is that these are three different businesses inside your business, and running them with one process guarantees mediocrity in all three.

Why do referrals convert multiples better than any other source?

The data on Seeds is unambiguous. In NAR's 2024 Profile of Home Buyers and Sellers, 40 percent of buyers found their agent through a referral from a friend, neighbor, or relative, and another 21 percent used an agent they had worked with before. On the listing side, 66 percent of sellers found their agent through a referral or used an agent they had worked with in the past, per NAR (2024). More than half of the entire market, in other words, never really shops for an agent. It asks someone it trusts and takes the answer.

Conversion economics follow the same pattern. Referral leads convert roughly 30 percent better than leads generated by other marketing channels, per referral marketing research compiled by GrowSurf (2026). A study of referral programs published by Wharton researchers Schmitt, Skiera, and Van den Bulte found referred customers carry about a 16 percent higher lifetime value and materially higher margins than non-referred customers (2011). And the effect compounds with tenure: in NAR's 2025 Member Profile, agents with 16 or more years of experience reported that repeat business accounted for 49 percent of their volume and referrals another 32 percent, against essentially zero for brand-new agents.

Trust is the mechanism. A referred luxury buyer skips the evaluation phase almost entirely. He arrives with borrowed conviction: someone whose judgment he already trusts has vouched for you, which means your first conversation starts where a cold lead's fifth conversation ends. In a $5 million transaction, where the buyer's real fear is not price but embarrassment and misplaced trust, that borrowed conviction is worth more than any ad budget.

A referral is a lead that arrives with the trust already built. Everything else you generate must earn in weeks what a referral inherits in one sentence.

If Seeds are so good, why can't they be the whole plan?

Because you cannot manufacture them on demand, and a business you cannot throttle is a business you do not control. Seeds are a lagging indicator of past excellence: this year's referrals were earned two, five, ten years ago. That creates three structural problems. First, feast and famine: referral flow is lumpy, and a slow quarter cannot be fixed by asking harder. Second, ceiling: your referral volume is bounded by the size and activity of your sphere, which is why even elite agents plateau. Third, fragility: a market shift, a move, or a single lost anchor relationship can cut the flow overnight, with no lever to compensate.

The answer is not to de-prioritize Seeds. It is to systematize them, most teams run referrals on memory and good intentions, with no defined ask, no tracking, and no reciprocity engine, and then to add engines you can actually control. NAR's member data shows the median agent gets 28 percent of business from repeat clients and 22 percent from referrals, per NAR (2025), which means roughly half of the typical book is uncontrolled. The goal of the mix is to keep that half performing while building the other half deliberately.

Why do inbound and outbound need completely separate workflows?

Because the buyer's psychological state is opposite in each. A Net lead raised her hand: she downloaded the market report, commented on the reel, filled the form. She expects speed and relevance. The defining metric is response time, and the research is brutal about it: contacting a web lead within 5 minutes makes you about 21 times more likely to qualify that lead than waiting 30 minutes, per the Lead Response Management study by Dr. James Oldroyd with InsideSales.com (2007). Inbound is a race against decay. The workflow that wins is instant response, fast qualification, and structured nurture for the majority who are twelve months from transacting.

A Spear target did not raise his hand. He was chosen. He expects nothing from you, which means the workflow is the inverse: research before contact, a specific reason for the outreach, personalization that proves the message could only have been written for him, and a cadence of multiple touches over weeks. Speed is irrelevant; persistence and precision are everything. Ross's insight at Salesforce was that these two motions are so different that they should not even be run by the same person, which is why Predictable Revenue (2011) argues for role specialization: dedicated prospectors who open conversations, separate from closers who run them to contract.

When teams run both lead types through one undifferentiated process, both die. Inbound leads get outbound treatment, slow, heavy, salesy, and go cold before contact. Outbound targets get inbound treatment, a generic drip built for hand-raisers, and mark it as spam. Separate sources, separate metrics, separate cadences, often separate people. That is the operational meaning of the framework.

What does the 20/20/60 mix look like in practice?

The mix we consider healthy for a luxury team building toward predictability is 20 percent Seeds, 20 percent Nets, and 60 percent Spears, measured by share of new pipeline created, not by closed volume. The number surprises people, because it inverts the industry's actual distribution. It is deliberate: pipeline you create through directed effort is the only pipeline that scales on command, so the system is designed to over-weight the controllable engine while the uncontrollable ones keep compounding in the background.

• Seeds at 20 percent of new pipeline: systematized, not passive. Quarterly touch plans for past clients, a defined referral ask after every closing and every milestone, tracked introductions, and reciprocity with the advisory bench, attorneys, wealth managers, immigration counsel, who serve the same client.
• Nets at 20 percent: founder-led content and paid campaigns that generate hand-raisers, wired to an instant response system, because inbound without a five-minute speed layer is a budget bonfire. Every Net lead lands in a nurture architecture, not an inbox.
• Spears at 60 percent: named-target prospecting. Expired and withdrawn luxury listings, owner lists in specific buildings, developer relationships, geographic farms, and strategic partners, worked in multi-touch cadences with weekly activity quotas that make next quarter's pipeline a math problem instead of a hope.

Note what the mix implies about closed business. Because Seeds convert at multiples of the other engines, a 20/20/60 pipeline mix can still produce closed volume that is one-third or more referral-driven. You are not abandoning the best source. You are refusing to let the best source be the only source.

The mix also changes how you read a slow month. In a referral-only business, a slow month is a mystery and a source of quiet panic, because the input that produced it happened years ago and cannot be adjusted. In a 20/20/60 business, a slow month is a diagnostic readout. If Spears activity held quota and pipeline still dipped, the message or the target list needs work. If Nets volume fell, a channel or a campaign decayed. If Seeds went quiet, the touch plan slipped. Each engine has its own dashboard, its own leading indicators, and its own repair procedure. That is the practical difference between running a business and riding one, and it is the entire reason the framework exists.

How do you migrate from a referral-only book to a mixed pipeline?

In sequence, not all at once. The first move is measurement: tag every opportunity in the CRM by source engine, because you cannot manage a mix you cannot see. Most teams discover their true distribution is something like 70/25/5, heavily Seeds and accidental Nets, with outbound near zero. The second move is protecting Seeds with a system, the touch plans and referral asks above, which usually produces a quick lift because the asset was underworked. The third move is standing up the Spears engine: pick two named-target segments, write the cadences, set weekly attempt quotas, and staff it, either with dedicated hours or a dedicated person. The fourth move is pointing Nets at the same segments so content warms the market outbound is hunting.

This is a build project with phases and deadlines, the kind of 30 to 90 day construction we run at Growth Ignis, structured so that Predictability comes first, then Profitability, then Freedom for the founder. A team that completes it stops asking the January question that haunts referral-only businesses, where will this year's deals come from, because the answer is written in the activity math: this many spear attempts, this many net leads, this many seed touches, producing a target of 10 to 15 qualified opportunities per week.

Aaron Ross's framework endures because it converts a mystical thing, momentum, into an engineering diagram. Referrals are a gift. Inbound is a garden. Outbound is a machine. Predictability begins the day you stop treating all three as one thing called leads.

FAQ

• Why 60 percent outbound if referrals convert so much better? Because the mix measures pipeline creation, not closed deals. Outbound is the only engine that responds immediately and proportionally to effort, so over-weighting it gives you control. Referrals still punch far above their pipeline share in closed volume because of their superior conversion.
• Does outbound even work in luxury real estate? Yes, when it is precise. Luxury outbound is not cold-calling strangers at dinner. It is researched, named-target outreach to expired listings, specific buildings, and developer relationships, with a specific reason for contact. Predictable Revenue calls this Cold Calling 2.0: prospecting without cold calls.
• How long before the mix changes my closed volume? Expect the Seeds system to lift results within one quarter, Spears within one to two quarters as cadences mature, and Nets to compound over two to four quarters. The sequencing exists precisely so the fast engines fund patience for the slow ones.