Apollo

Apollo exposes a broad set of objects: contacts, accounts, deals (opportunities), outreach emails, calls, sequences, tasks, and users. This integration focuses on three object types — outreach emails, calls, and deals — which together cover the full span of the outbound sales motion, from initial prospect engagement to closed revenue. Contacts, accounts, and sequences were not included: they describe the prospecting database and workflow configuration, not individual performance output. Seven KPIs were retained against three criteria: ability to attribute to an owner, resistance to gaming, and balance between leading and lagging indicators.

Email outreach: activity, outcome, and the discipline to pair them

Three KPIs cover outbound email. Emails Sent counts successfully delivered messages. Positive Replies counts responses classified as expressing willingness to meet — a signal generated by Apollo's reply classification engine, not a raw reply count. Bounce Rate measures the proportion of sent emails that were rejected before delivery.

The three indicators are designed to be read together, and the design reflects a specific gaming risk. Outbound email volume is the single most gameable metric in sales engagement: an SDR under pressure to show activity can send a high volume of poorly targeted messages to low-quality contacts. Read alone, Emails Sent reveals nothing about the quality of the effort. Positive Replies corrects part of this: an interested response requires relevance in both targeting and messaging, and cannot be manufactured by volume alone. Bounce Rate corrects the remaining part: systematically targeting low-quality or purchased contact lists produces a higher bounce rate, making the volume inflation visible. The combination of the three closes the gaming loop.

A tension also exists between Positive Replies and Bounce Rate that is analytically distinct from the gaming consideration. A low Positive Replies rate combined with a low Bounce Rate may indicate that the list quality is adequate but the messaging or targeting is wrong. A high Bounce Rate combined with a low Positive Replies rate points to a list quality problem upstream. Distinguishing between these two configurations changes the diagnosis and the corrective action.

Phone outreach: connected calls as the only defensible unit

One KPI covers calls. Calls Connected counts outbound calls that reached a live person, not total dials placed. This distinction is deliberate. Total dials is the outbound calling equivalent of emails sent: it measures button presses, not conversations. A representative who dials the same number repeatedly, or systematically calls numbers known to be inactive, accumulates dial counts without producing any commercial value. Calls Connected requires a human to answer, which imposes a floor of reality on the metric that dials does not have.

Calls Connected, like Emails Sent, is a leading indicator: it measures activity that precedes outcomes rather than outcomes themselves. On its own it does not distinguish between a call that generated a meeting and a call that ended in a brief hang-up. For teams tracking call-to-meeting conversion, Calls Connected provides the denominator; the numerator would need to be tracked in a downstream object such as deals. This integration does not currently include a call outcome KPI, as Apollo's call data does not expose a standardized outcome field with sufficient coverage across plans.

Deal pipeline: leading creation against lagging conversion

Three KPIs cover opportunities. Deals Created counts new pipeline entries per owner. Won Amount sums the revenue of closed-won opportunities. Win Rate measures the proportion of closed opportunities that were won, applied on a rolling basis.

Deals Created is the only leading indicator in this group and follows the same gaming logic as Emails Sent: an AE under pressure to show pipeline activity can create opportunities of low credibility. Won Amount and Win Rate together act as the downstream corrective. Won Amount is immune to pipeline inflation because it only registers upon a closed-won event; Win Rate degrades mechanically if the created pipeline lacks real qualification, since the denominator grows faster than the numerator. A sustained increase in Deals Created combined with a declining Win Rate is therefore a diagnostic signal of qualification deterioration, not a sign of improving commercial momentum.

Won Amount and Win Rate carry a different tension between them that is independent of the gaming question. A high Won Amount combined with a low Win Rate may indicate that the team is selectively closing a small number of large deals while losing many smaller ones — a pattern that produces good revenue numbers on a given period but concentrates risk. A high Win Rate combined with a modest Won Amount may indicate strong qualification discipline but limited pipeline ambition. These two configurations require opposite management responses, and Win Rate alone does not surface the difference.

Scope and limits of the integration

Apollo records email delivery status, call connection status, and deal stage transitions, but it does not measure the quality of a conversation. A connected call may last thirty seconds or thirty minutes; the API does not distinguish between the two at the KPI level used here. A positive reply may come from a contact who quickly disengages after the first meeting; the metric captures the intent signal, not the downstream outcome. Won Amount is the single metric in this integration that measures a real commercial result without ambiguity.

This integration also depends on accurate user attribution throughout the Apollo workflow. Emails and calls are attributed by the user who sent or placed them; deals are attributed to their owner in Apollo. If ownership is not maintained as representatives change or deals are reassigned, KPI values will misalign with actual individual contribution. The reliability of these indicators depends directly on the team's discipline in maintaining accurate records in Apollo.