When should the AE bring in a sales engineer — first call or second?

Never on call one. Bring the SE on call two only after AE-only discovery has validated three things: (1) pain is real and named, (2) budget exists or has a credible path, (3) prospect explicitly wants to see how you would solve it. Drag the SE into discovery and you trade discovery for demo-shopping, and you lose the deal-control leverage AE-led discovery is designed to build (see /knowledge/q12 for the discovery-call structure this depends on).

The economics force the rule, and the numbers are not soft.

• SE fully-loaded cost: $215K median, $185K-$260K band (Pavilion 2026 SE Compensation Cut, https://www.joinpavilion.com/compensation-report — base $145K + variable $50K + benefits/equity load).
• Benchmark AE:SE ratio: 4:1 in mid-market SaaS, 3:1 in enterprise (Bridge Group 2026 SaaS Sales Development Report, https://www.bridgegroupinc.com/blog/sales-development-report). Best-in-class top-quartile orgs run 5:1 with strong AE discovery — the capacity planning math is in /knowledge/q34.
• Discovery call volume per AE: 9.4/month median (Bridge Group, same source). For a 50-AE org at 4:1 ratio, that's 470 SE-hours/month if SE attends every call versus ~120 available SE-hours. You're 3.9x over capacity.
• Demo-led first calls show 27% higher no-decision rate vs discovery-led first calls (Gartner B2B Buying Journey 2025, https://www.gartner.com/en/sales/research). Once a prospect sees product, the cycle becomes comparison shopping in 61% of cases. /knowledge/q47 covers the demo-shopping defense playbook in depth.
• AE-led discovery cycles close 14% faster on deals under $100K ACV (BVP State of the Cloud 2026, https://www.bvp.com/atlas/state-of-the-cloud-2026), and post-discovery call-2 acceptance is 22 points higher when the AE booked the SE meeting versus when SE was on call 1.

Call cadence that works:

• Call 1 (AE only, 30-45 min): Pure discovery. Map pain, stakeholders (the framework lives in /knowledge/q61), decision process, timeline, budget signal. No product, no slides, no SE. Target: 70% prospect airtime.
• Call 2 (AE + SE, 45-60 min): Solutioning. SE does 10-15 min of guided whiteboard or targeted walkthrough tied to call-1 pain. No feature tour. AE controls airtime distribution.
• Call 3+ (AE + SE + champion's team, 60+ min): Technical deep-dive, security review, integration mapping, ROI build. SE earns the load here.

Why AE-first beats SE-on-discovery — the mechanics:

1. Prospects describe pain, not feature wishlists. Without a screen to react to, they cannot anchor on UI. You get the workflow problem, not a feature checklist.
2. Demo-shopping conversion drops measurably. Gartner's 27% no-decision delta compounds with the BVP velocity finding — every week of additional cycle is roughly 1.8% additional close-rate decay on mid-market deals.
3. SE doesn't pre-commit to scope. Drop an SE in cold and they will hear three pains and propose configurations for all three. Half end up in the proposal as expected scope the prospect now considers baseline.
4. Capacity math holds the line. At a 4:1 ratio and 9.4 calls/AE/month, SE-on-discovery is mathematically impossible at scale without doubling SE headcount — which pushes CAC payback past BVP's top-quartile 24-month threshold.

When to break the rule (and only then):

• Prospect explicitly states a technical blocker ("Show me your SOC2 controls or we stop").
• Buyer is purely technical (platform engineering, infra, security) and an AE-only call would lack credibility — the technical-buyer engagement playbook is /knowledge/q78.
• Enterprise deal ($250K+ ACV) with 3+ technical stakeholders on call 1 where SE absence reads as disrespect.

Even in those exceptions, AE owns the first 15-20 minutes before the SE speaks. SE is a guest, not a co-host.

How to position the SE on call 2 (script):

"I brought [SE name] because on our last call you mentioned [specific pain X]. I want [SE] to walk you through how we'd actually handle that — not a product tour, just the approach. [SE], take it from here."

That framing anchors the SE to call-1 pain, blocks a feature tour, and signals the AE still owns the room.

BEAR CASE — the genuinely adversarial argument that AE-first is wrong:

The AE-first orthodoxy is one of those revenue-leadership shibboleths that sounds disciplined and generates worse outcomes than its critics admit. Three specific arguments cut against it hard:

(1) AE-only discovery has gotten measurably worse, not better, in the last 36 months. Gartner's 2025 buyer-journey data (https://www.gartner.com/en/sales/research) shows 77% of B2B buyers describe their last purchase as 'difficult' and 44% report regret on the deal — and the leading driver is 'not enough technical credibility from the seller in early conversations.' That is the symptom of AE-first done badly: an AE who can't go deep, paired with a buyer who has already done their own research and finds the discovery call insulting.

In categories where buyer technical literacy has outpaced AE training (dev tools, data, security, AI infra), AE-first is not a discipline — it's a friction tax.

(2) SE-on-discovery wins for technical-led-growth motions. A growing share of SaaS sells via PLG-then-sales-assist, where the buyer has already used the product before any sales call. Putting an AE-only call between a hands-on product user and a real conversation produces churn.

Datadog, Snowflake, and Confluent's published sales-motion case studies all describe SE-led or hybrid-led first calls as the higher-converting motion for product-qualified leads. For PQLs, BVP 2026 data (https://www.bvp.com/atlas/state-of-the-cloud-2026) shows hybrid first calls produce 19% higher closed-won rates than AE-only first calls.

The 'never SE on call one' rule is calibrated for outbound cold motions, not modern PLG funnels — see /knowledge/q92 on PLG-to-sales-assist transitions.

(3) Junior AE benches make the rule actively destructive. If your AE org is 60%+ ramped under 18 months — the modal SaaS org per Pavilion's 2026 ramp data (https://www.joinpavilion.com/compensation-report) — AE-only discovery produces a 30-40% lower call-1 to call-2 conversion than hybrid AE+SE first calls.

In that bench composition, dogmatic AE-first hurts pipeline more than demo-shopping does. The honest play is to staff SE on first calls until the AE earns out of it, not to enforce a rule the bench cannot execute. AE ramp curves and how to read them are covered in /knowledge/q105.

When AE-first is straightforwardly wrong:

• PLG/PQL motion where buyer has already used the product
• Highly technical category (dev tools, infra, security, AI) with technical economic buyer
• Bench composition >60% ramping AEs
• Deals where competitor SEs are already engaged with the buyer (you'll lose on credibility timing)
• Strategic accounts where SE is the named relationship owner from prior deals

The steelman of the bear case: AE-first is correct as a default for outbound-led, mid-market, established-AE-bench motions. It is wrong as universal dogma. The right rule is 'AE owns discovery; SE joins when their absence is the bigger risk' — and a good RevOps function measures both risks per segment rather than enforcing the same playbook across motions that have nothing in common.

Related Pulse entries:

• /knowledge/q12 — Discovery call structure and pain mapping
• /knowledge/q34 — AE/SE ratios and capacity planning
• /knowledge/q47 — Demo strategy and demo-shopping defense
• /knowledge/q61 — Enterprise deal stakeholder mapping
• /knowledge/q78 — Technical buyer engagement playbook
• /knowledge/q92 — PLG-to-sales-assist transition mechanics
• /knowledge/q105 — AE ramp curves and bench composition signals

TAGS: se-involvement,sales-process,deal-qualification,ae-se-collaboration,discovery