How Buy-Side Firms Handle Expert-Network Call Overruns: 7 Structural Models for Billing, Consent, and Compliance

Expert-network calls are almost universally scheduled and priced in 60-minute increments. The default hourly model runs across GLG, Guidepoint, AlphaSights, Third Bridge, Dialectica, and Coleman, with standard experts billed at roughly $500 to $1,500 per hour and specialists higher. Analyst-led calls, however, frequently overrun as the conversation gets valuable, and how firms handle those extensions is a structural question that touches billing, compliance, recording consent, and analyst behavior at once.

Below are seven distinct models currently in practice across long-only, hedge-fund, and multi-manager buy-side desks. None is universal, and larger platforms often blend two or three depending on the pod, the sector, and whether a compliance chaperone is on the line.

1. Hard-Stop at 60 Minutes

The simplest model: when the clock hits 60 minutes, the account manager or the platform's call software terminates the session. No overrun, no exceptions, no side-channel to extend.

This is the default at cost-controlled long-only shops where budget discipline outweighs the marginal insight of an extra 10 minutes. It is also common at firms where junior analysts run most of the calls and where the research head has decided that unbounded expert time is not a good use of the research budget. At standard hourly rates in the $500 to $1,500 range, an analyst who routinely runs 20 minutes over across 15 calls a month is spending an extra $2,500 to $7,500 that was never approved.

The operational tradeoff is real. Analysts complain that the best material often comes in the final 10 minutes, once the expert has warmed up and started volunteering specifics. Firms that run this model accept that cost as the price of predictable spend.

2. Pre-Authorized Overrun Buffer

A middle path: the analyst is pre-approved to extend the call by a defined buffer, typically 15 or 30 minutes, without needing to ping anyone. The overrun is billed pro-rata or at a half-hour increment depending on the network's terms.

This model treats the overrun as a routine operational reality rather than an exception. It works best when the expert budget is tracked at the analyst level and the analyst has both the authority and the accountability to spend it. The buffer creates a soft ceiling that keeps calls from stretching to two hours while removing the friction of a real-time approval loop.

The compliance angle here is that the buffer is pre-scoped. The expert agreed to the extended window at the top of the call, or the network's terms of engagement specify that hourly bookings include a standard courtesy extension. Nothing new happens at minute 61.

3. Real-Time PM Approval

At multi-manager pods, expert budgets typically sit with the portfolio manager rather than with a central research function. When a call runs long, the analyst pings the PM (Slack, Teams, Bloomberg chat) and gets a yes or no in real time.

This is the dominant model at Millennium, Citadel, and Point72-style platforms where each pod runs its own P&L and its own expert-network spend. The PM's approval is logged in the chat thread, which doubles as an audit trail. If the call goes another 40 minutes and the network bills for two hours, the pod's operations team can point to the timestamped approval when reconciling the invoice.

The workflow adds a small amount of friction, but PMs tend to prefer it. An expert call that has gone long usually means the analyst is onto something, and the PM would rather hear about it in real time than read a transcript the next morning.

4. Compliance Re-Consent Model

Some firms treat the moment the call passes its scheduled end as a new engagement window. Before the call continues, the expert is asked to reconfirm the compliance disclosures they made at the top of the hour: no new NDAs signed, no trading blackouts entered, no material events at their current or former employer that would change the scope of what they can discuss.

This matters most for calls with current operators, particularly in situations where a company has entered a quiet period or where the expert's employer has been subject to a corporate development event that could have introduced new restrictions since the call was booked. The re-consent creates a documented boundary between what the expert was cleared to say in hour one and what they are cleared to say in hour two.

The legal grounding for this is meaningful. In two-party-consent jurisdictions (California, Illinois, Massachusetts, Pennsylvania) and under GDPR, recording consent technically applies to the scope originally agreed. If the scope materially changes mid-call, a fresh confirmation is the defensible position.

5. Flat-Fee-Per-Call and Enterprise Subscription Structures

Firms on all-you-can-eat subscription models (the Tegus and AlphaSense enterprise tier, for example, which bundles transcript library access with a call allowance) sit in a different structural position. Overruns are not billed by the minute. The subscription price is fixed, and the fair-use provisions in the contract cap total consumption rather than per-call duration.

Under this model, the operational question at minute 60 shifts from "who authorizes the spend" to "is this call still worth the analyst's time." The billing pressure that shapes the hard-stop and buffer models does not exist, which changes analyst behavior. Calls tend to run longer on average at firms with enterprise subscriptions, and firms compensate with call-productivity metrics on the research side rather than budget caps.

The compliance and recording-consent questions still apply. The billing model does not change the underlying legal framework for what happens when a call passes its originally scoped duration.

6. Chaperoned-Call Hard Stop

When a call has a compliance chaperone on the line (a network-provided or in-house compliance officer whose job is to intervene if the conversation approaches MNPI), the chaperone typically terminates the call at 60 minutes regardless of analyst preference.

The reason is operational rather than doctrinal. Chaperones are scheduled in fixed hourly blocks, often stacked back-to-back across a compliance officer's day. Extending one call means the chaperone is late or absent for the next, which the network cannot support at scale. The hard stop is enforced at the chaperone's end, not the analyst's.

Firms that rely heavily on chaperoned calls (China-focused funds, healthcare desks with a lot of physician calls, any desk with elevated MNPI risk) build their research workflow around this constraint. If the topic requires more than an hour, they book two consecutive calls with a fresh compliance window between them rather than trying to extend a single session.

7. Recording-Boundary Segmentation

The most operationally sophisticated model handles the overrun at the recording layer. The call continues past 60 minutes, but the recording is segmented into two files: the scoped hour and the overrun.

The segmentation matters for downstream processing. Some firms flag the overrun portion for a separate compliance review before it enters the internal transcript library, on the theory that the risk profile of the extended conversation may differ from the scoped hour. Others exclude the overrun from AI-ingestion pipelines that feed internal research summarization tools, so that any MNPI-adjacent material that surfaced in the extended window does not propagate into automated outputs.

The segmentation is also useful for billing reconciliation with the network and for retention policy. If a firm's retention schedule treats scoped and out-of-scope material differently, a clean file boundary at minute 60 makes the policy enforceable rather than aspirational.

The Compliance History Behind These Structures

The reason overrun policy is codified at all traces back to the enforcement environment of the past 15 years. The Galleon and SAC-era prosecutions between 2010 and 2013 established that the content and duration of expert calls are evidence, and the SEC's 2014 settlement reinforced that firms are expected to have documented controls around expert-network engagement.

Overrun handling is where those controls meet operational reality. A call that runs 40 minutes long without a documented authorization, without a fresh consent check, and without a clean recording boundary is a set of unanswered questions in a future compliance review. The seven models above are the ways firms answer those questions in advance.