haggl.ai Blog

Segment Priority Isn’t Sort Order. It’s a Bet.

May 27, 2026|8 min read

Of all the fields on an ICP segment in haggl, priority looks the most boring. It’s a number. The UI shows segments sorted by it. Most vendors set it once during onboarding and never look at it again. That’s a mistake. Priority is doing more commercial work than almost any other field on the segment, and most of that work is invisible until you go looking for it.

This post is about what priority actually does in the protocol, what strategic question it answers, and the three or four common ways vendors set it wrong without realising.

The mechanic

When a buyer agent submits a negotiation, the payload optionally includes a target_segment_id — the agent’s best guess at which of the vendor’s ICP segments fits its principal. If the agent is confident, it passes the ID. If the offer comes back keyed to that segment, the agent moves on.

But agents often aren’t confident. The buyer might fit two segments. The agent might be a generic ChatGPT-style runtime with no prior knowledge of the vendor’s catalog. In that case target_segment_id is omitted, and the vendor’s negotiate endpoint has to decide which segment to score against.

It picks the highest-priority segment — the one with the lowest priority number. That’s the tiebreaker. The buyer is silently routed into the segment the vendor declared as their default, with that segment’s discount ceiling, that segment’s proof requirements, and that segment’s checkout link (see Right Vendor, Wrong Plan).

In practice, a meaningful share of agent sessions arrive without a target_segment_id. Across haggl design partners we see anywhere between 25% and 60% of sessions defaulting to the priority-1 segment, depending on how technical the buyer agent is and how legible the vendor’s segment names are. Priority decides what happens to those buyers.

The strategic question priority is answering

So the actual question priority is silently answering is:

When a buyer plausibly fits more than one of our segments and the agent doesn’t pick, which segment do we want them to land in?

Note what this is not asking. It’s not asking which segment is the biggest, which gives the deepest discount, or which one the marketing team likes most. It’s asking which cohort you’d rather close an ambiguous buyer into — given that the cohort choice determines unit economics, lock-in dynamics, and downstream sales motion.

Three concrete examples sharpen what that means in practice.

Example 1 — the returning customer who looks like a prospect

An Australian electricity retailer has two segments: Existing Customers (priority 1, 35% margin ceiling, deep retention discount authorized) and New Residential (priority 2, 18% margin). An agent shows up with a buyer who hasn’t logged into their existing account, can’t produce a bill from this retailer specifically, and reads to the agent as “a residential consumer in Melbourne.”

Both segments fit. The buyer probably fits New Residential more cleanly — the proof package looks like a new-buyer pattern. But there’s a non-zero chance this is an existing customer the agent didn’t cross-reference. With Existing Customers at priority 1, the default routing assumes retention until proven otherwise. The agent gets the retention package. If the buyer is genuinely new, they accept anyway (it’s a better offer). If they were existing, the vendor just defended a renewal that would otherwise have churned.

Flipping priorities — making New Residential priority 1 — reverses the bet. Now the default is “acquire,” not “retain.” You’ll close more cleanly-new buyers, but you’ll leak retention dollars on existing customers the agent didn’t recognise. Same segments, same ceilings, different default — materially different cohort economics.

Example 2 — the SaaS lane-choice

A developer-tools SaaS vendor has Enterprise Platform Teams (priority 1, 22% margin, multi-year contract motion, long-tail expansion) and Scale-up Engineering Teams (priority 2, 28% margin, faster sales cycle, NRR 135% historically).

A 200-engineer Series C company sits right on the boundary. Their agent could plausibly pick either. The vendor’s choice of priority encodes a strategic preference:

Enterprise first. Default-route ambiguous mid-market buyers into the enterprise lane. Slower close, smaller initial deal, but you’re betting on the multi-year contract sticking and the procurement process selecting for stable buyers.
Scale-up first. Default-route the same buyer into the scale-up lane. Faster close, deeper initial discount, optimised for landing the seat-count expansion.

Same buyer, same product, same protocol — two completely different commercial bets, driven by which integer is smaller.

Example 3 — the multiplier vs the direct customer

A coworking vendor has Remote Teams (priority 1, 18% margin, direct spend) and Digital Agencies (priority 2, 10% margin, lower direct LTV but they bring clients onto the same workspace).

An agency-run remote team fits both. Priority 1 = Remote Teams means “we’d rather treat them as a direct customer with bigger initial spend.” Priority 1 = Digital Agencies means “we’d rather treat them as a multiplier — smaller direct ticket, but we get their five clients next month.” Network effects play vs direct-revenue play, decided silently by priority order.

The mistakes vendors make

Four patterns we see consistently when reviewing vendor configurations.

Setting priority 1 = highest discount. The instinct is “our most generous offer should be our default.” This conflates priority (strategic preference) with max_margin_percent (discount ceiling). They’re independent. You can prioritise a low-discount segment because that’s the cohort you most want to grow, and authorise deeper discounts on a non-priority-1 segment to handle edge cases.
Setting priority by volume. “Most of our deals come through the New Residential segment, so put it first.” This conflates priority with frequency. New Residential is already the segment most buyer agents will explicitly target. The one that gets the priority benefit is the segment buyer agents don’t target — the ambiguous ones — which is usually the one your business needs more of, not the one you already get the most of.
Never adjusting priority. The right priority order in your land-grab phase isn’t the right priority order in your retention phase. Most vendors set it on day one and forget. Plan to revisit quarterly — it’s a 30-second change in the dashboard, with effects on the next agent session.
Setting priorities that contradict the segment intel. If a segment’s intel.blurb describes it as “a small but high-network-value cohort,” priority 1 is probably wrong — you don’t want it to be the default destination for ambiguous buyers, you want it reserved for buyers who actually fit. Priority and intel should agree on what each segment is for.

The dynamic strategy

Priority order should change with the business’s posture. A useful three-phase rule of thumb:

Land-grab phase. Priority 1 = the segment that gives you the most pipeline velocity, even at thinner margins. You’re paying for cohort breadth. Acquisition-skewed segments win.
Margin-protection phase. Priority 1 = the highest-LTV segment, usually retention or existing-customers. You’re paying for cohort depth. Buyer agents that can’t prove segment-specific fit default into the deepest-LTV bucket.
Network-effects phase. Priority 1 = the multiplier segment. Smaller direct deal but each closed buyer brings 3–5 more. You’re paying for coefficient of expansion.

Most vendors cycle through these as they grow. The priority order should cycle with them.

What this connects to

Priority is upstream of nearly every other lever on the segment. The cohort it routes ambiguous buyers into is the cohort the discount ceiling applies to (see How Much Discount Should You Authorize?), the cohort the proof requirements apply to (see Not Every Segment Needs Proof), and the cohort the checkout URL points at (see Right Vendor, Wrong Plan). Get the priority wrong and the rest of the segment design starts working on the wrong cohort.

The point

Priority looks like a sort order. It is, technically. But the sort it’s ordering is the order in which the vendor would prefer to close ambiguous buyers. That makes it one of the most consequential commercial declarations on the segment definition, and it deserves more than five seconds of thought during onboarding.

Set it once, watch the negotiation logs for which cohort the ambiguous-buyer bucket actually lands in, and adjust quarterly as your posture changes. The lever is quiet but it moves real dollars.

Related Reading: