Agency awards and commercial evidence: what to evaluate

Inspect what recognition measures

Ask for category, date, criteria and submitted work. Creative execution, culture and partnership awards address different questions from account selection, sales coordination or incremental pipeline. Treat recognition as specific information.

No head-to-head Outsell-versus-The-Smarketers experiment was supplied. A polished website or AI label cannot establish delivery effectiveness either. Apply the same standard: named work, outcome, baseline and limits.

Use the same account brief

Ask shortlisted providers for a sourced hypothesis, two role-specific messages, an execution plan and a measurement definition against the same fictional or permissioned brief. Compare reasoning and usability before polish.

Check unsupported intent, confused entities and claims that a named person performed anonymous activity. Ask how missing evidence is handled. Do not reward confident guesswork or request private customer information.

Observational methods often did not recover the effects found by randomised experiments. Consumer advertising on one platform. This informs measurement design, not the effect size of an enterprise ABM programme. Brett Gordon, Florian Zettelmeyer, Neha Bhargava and Dan Chapsky (2019): A Comparison of Approaches to Advertising Measurement.

The practical workflow

1. Inspect recognition criteria
2. Use one comparable account brief
3. Review reasoning and sources
4. Confirm delivery ownership
5. Agree a measurable pilot

Compare the approaches

Inspect the operating model

Identify the owners of account records, claim approval, handoff and failed integrations. For agentic work, inspect permissions, traces, acceptance checks and retry limits. For specialist-led work, inspect coordination.

Providers can combine these methods. The Smarketers archive includes AI content, tools and ABM cases, so describing it as purely manual would be inaccurate. Choose demonstrated fit rather than a convenient competitor label.

Read case studies carefully

Check starting condition, timeframe, channels, budget and attribution. Distinguish qualified opportunities, booked meetings, attended meetings, sourced pipeline and influenced pipeline.

The measurement research explains why observed outcomes and incremental effects can differ. It does not invalidate every case. It identifies questions to answer before transferring an impressive result to another programme.

The paper examines how promotional language presents AI agency and human roles in learning. It is not an agency performance experiment and provides no Outsell-versus-competitor or 2× result. Isabel Pedersen and Ann Hill Duin (2022): AI Agents, Humans and Untangling the Marketing of Artificial Intelligence in Learning Environments.

Make a revisable procurement decision

Score account expertise, capacity, evidence quality, data access, coordination and terms. Include transition costs and client review responsibilities. Agree continue, revise and stop criteria.

Outsell proposes software-supported account work with traceable evidence and defined quality gates. Whether this is the better enterprise fit must be established through evaluation.

Your next-action checklist

• Awards: Inspect category and submitted work. Check the limitation: criteria may not match the need.
• Public case: Request definitions and context. Check the limitation: may lack a control.
• Sample account play: Check source support. Check the limitation: sample is not production delivery.
• Pilot: Agree review criteria. Check the limitation: sample and transfer limits.

Use the comparison to choose a bounded next step. Record the evidence, the responsible owner, and the review decision before extending the play to additional accounts.

Questions this guide answers