Loading Outsell AITHE OUTSELL LEARNING LIBRARY
Clear explanations of agentic marketing and account-based marketing. Learn what the evidence says, where it applies and how to test it in your programme.
18 research-led guides · Worked examples · Ungated worksheets · English, French and German
Choose the path that matches your current work. Begin with the agent research path when you need reliable retrieval and source support. The buyer-value path helps you choose useful personalisation. The measurement path addresses prediction, rare events and causal comparisons. The operating-model path connects technology to account ownership and review responsibilities.
Each guide explains the studied question, the finding and its boundary before suggesting an account application. Follow the worked example, then use the ungated worksheet to record your own input, decision and acceptance criteria. A guide is most useful when it produces one artefact your team can review, rather than a list of tools to buy.
When reviewing a source, ask what task was measured, who participated, what the comparison was and whether the outcome matches your proposed claim. A productivity finding can inform a workflow test without establishing an ABM revenue effect. Keep the source finding and your proposed application separate in the worksheet.
6 CONNECTED GUIDES
Understand retrieval, citations, context and safe tool use before connecting account research to activation.
01 / Account intelligence
Retrieve relevant, dated passages before drafting. Preserve the identity and source behind each assertion, and allow the workflow to return an unresolved question.
02 / Agent operations
Check whether the linked source supports the exact assertion. Citation presence, factual correctness and complete support are separate quality questions.
03 / Agent operations
A large context window is capacity, not a guarantee of reliable evidence use. Test critical facts and constraints in different positions and prefer a focused evidence packet.
04 / Agent operations
Give the agent a clear question, narrow tools and a verifiable stopping condition. Evaluate completed account work, including failed actions and review.
05 / Agent operations
External text can contain instructions designed to redirect an agent. Treat retrieved content as data and enforce permitted actions outside the model.
06 / Account intelligence
Record the original entity, the transformation and the responsible owner. A final summary alone cannot explain how a source became a commercial recommendation.
5 CONNECTED GUIDES
Choose the right depth, preserve trust and give each stakeholder evidence for the decision they face.
01 / Buying journeys
Evaluate correctness, role relevance and total production effort before measuring responses. Faster writing does not establish incremental pipeline.
02 / Buying journeys
Personalise around a relevant business decision and explainable evidence. Avoid implying access to private behaviour or a person’s intentions when you have only account-level signals.
03 / Buying journeys
Increase specificity when the buyer’s requirements and evidence justify it. Early exploration, solution evaluation and a named-account business case need different content.
04 / Buying journeys
Map the question, evidence and next action for each buying task. Keep those tasks connected across stakeholders and touchpoints instead of producing a large sequence of generic assets.
05 / Buying journeys
Preserve the business question, evidence and agreed next action across touchpoints. Choose channels from observed buyer needs rather than copying an average channel count.
4 CONNECTED GUIDES
Understand probability, rare conversions, holdouts and the difference between predicted response and causal impact.
01 / Measurement
A ranking score and a calibrated probability serve different purposes. Define the event and window, then compare predicted probabilities with later observed outcomes.
02 / Measurement
When conversions are uncommon, overall accuracy can hide missed opportunities. Report precision and recall at the team’s capacity, alongside prevalence and the cost of errors.
03 / Measurement
Preserve eligibility and assignment before interpreting exposure. Attribution records a relationship between touches and outcomes; a credible experiment addresses what would have happened without the intervention.
04 / Measurement
Propensity predicts an outcome under observed conditions. Uplift asks how an intervention changes that outcome; it needs a credible comparison and cannot be inferred from engagement alone.
3 CONNECTED GUIDES
Design complementary work, relationship capability and observable operating checks for an enterprise programme.
01 / Agent operations
Compare complete task designs. Human involvement helps when it supplies useful judgment; an undefined review step can add time without correcting the right errors.
02 / Account strategy
ABM coordinates account-focused marketing and sales work. Key account management adds continuing relationship and delivery responsibilities; the two need shared objectives and clear ownership.
03 / Agent operations
Connect each material failure mode to a check, an accountable owner and a recovery action. Evaluate actual workflow behaviour rather than treating a policy document as proof of reliability.
A controlled experiment, a buyer survey and a technical specification answer different questions. The guides explain the relevant finding before linking to the original publication.
Worked examples are fictional teaching scenarios. Applications to ABM are proposed methods to test, rather than reported client results.
Inspect the research libraryYour privacy choices
We use essential browser storage to remember this decision. We do not use analytics, advertising, or tracking cookies.
If optional cookies are added later, they will load only if you choose Accept. Essential storage remains on either choice.