Less friction. More useful work.

Marketing has grown faster than the systems around it. Requests arrive from everywhere, briefs are incomplete, approvals stall, reporting takes days of manual assembly, and the same corrections are made every month.

AI has been tried. Licences were bought, prompts were shared, a few people became fluent. The team is no faster, the output is no more consistent, and nobody can point to a workflow that improved.

• A marketing team of any size whose work is managed through messages, spreadsheets and memory.
• A business that has adopted a work-management tool without the processes to make it useful.
• A content or campaign operation where review and approval, not production, is the bottleneck.
• Leadership that wants AI applied to real operational work, with a measured result rather than a demonstration.
• A function preparing to scale output or headcount and wanting the operating model settled first.

• A map of how work moves today, where it waits and what it costs in effort.
• An operating model: roles, ownership, approval paths and standard procedures for campaigns and content.
• A configured or improved work-management system that reflects that model.
• One working automation or AI-assisted workflow, tested with the team, documented and handed over. AI is described by the task it supports and the result being measured: for example, first-draft briefs assembled from a structured intake, or weekly reporting compiled from agreed sources for human review.
• A measurement and adoption plan with a baseline, a review date and a named owner.

• Access to the people who request, produce and approve marketing work.
• Administrative access to the tools in scope, or a colleague who holds it.
• Agreement on one workflow to automate first, chosen for frequency and measurable effect rather than novelty.

Changes are measured against an agreed baseline captured before the work begins: cycle time from request to delivery, manual effort per item, error or rework rates, adoption of the new procedure, or operating cost.

The automated workflow has its own acceptance tests and a review period in which results are checked by a person before the workflow is relied upon.

Three kinds of work are kept distinct. Process design and operating procedures are delivered directly. Configuration of work-management and automation tools is delivered directly within the agreed scope. Specialist software engineering, custom integrations and data pipelines are identified, specified and coordinated with appropriate specialists, not delivered as part of this engagement.

Automation is limited to the workflow agreed, with review controls in place. Nothing is set to act without a defined check, and nothing is connected to systems outside the agreed access.

Tooling recommendations are made on fit, not on partnership. No software is advertised or resold.

Relevant career examples are shared in conversation, with the organisation through which the work was delivered named and the lead practitioner's contribution stated. Engagements commissioned through INOKULUM are documented as they complete, with the client's consent. No case studies, testimonials or client names are published without that attribution.