In short
- AI is moving from pilots to scale, but results lag: in McKinsey’s global survey 44 % scale AI across the enterprise and 37 % see any EBIT impact; in Gartner’s CMO survey 70 % say their marketing processes are not mature enough to scale it.
- Automation earns trust step by step: first the AI that watches and reports, then the AI that drafts, last the AI that acts.
- In Acontema every level has its scheduled work, watching agents check that it happened, and each kind of automated task has named people who answer for it.
1Automation is the challenge, not only the answer
Worldwide, companies are moving from trying AI to running it. In McKinsey’s global survey, 44 % of respondents say AI is scaling across their enterprise, up from 38 % a year earlier, and among organisations with more than $1 billion in revenue 40 % scale AI agents in at least one function, up from 27 %. The results have not followed: the share reporting any EBIT impact from AI is essentially unchanged at 37 %, and only about 6 % count as high performers.(McKinsey, The state of AI in 2026, Key takeaways; “Use of AI is deepening as organizations move beyond experimentation”)
Marketing shows the same gap between ambition and readiness. In Gartner’s 2026 CMO Spend Survey, marketing leaders put on average 15.3 % of their budget into AI, and 70 % call becoming an AI leader a critical goal for 2026. Just as many say their marketing processes are not mature enough to implement and scale AI, and only 30 % report mature AI readiness, while budgets stay flat at 7.8 % of company revenue.(Gartner CMO Spend Survey 2026, press release, “CMOs Increase AI Investment, But Readiness Lags”)
AI ambition and readiness in marketing
Share of CMOs and marketing leaders; North America, United Kingdom and Europe; n = 401, January to March 2026
- Becoming an AI leader is a critical goal for 202670 %
- Marketing processes not mature enough to implement and scale AI70 %
- Lack the budget to deliver the 2026 strategy56 %
- Insufficient resources54 %
- Mature or fully developed AI readiness30 %
The ambition is there and the money is moving; the processes that would carry it are not.
The risk is that CMOs invest in AI tools faster than they build the data foundations, processes, governance and talent required to scale them.
What separates the few who see results is not the budget alone. Nearly three-quarters of McKinsey’s high performers have fundamentally redesigned workflows because of AI, against one-quarter of other respondents, and the authors name human-in-the-loop design, impact measurement and risk management among the practices that keep deployments running.(McKinsey, The state of AI in 2026, “High-performing organizations are transforming with AI”; commentary by Tara Balakrishnan)
A European survey shows the same picture from inside marketing teams. Bitkom, the German digital association, asked 180 companies of the digital industry: automating marketing processes is named as an internal challenge as often as cost pressure, and about a third miss an AI strategy or say AI is not integrated well enough. The sample is small and not representative, but it matches the global pattern.(Bitkom, Marketing im digitalen Wandel 2026, pp. 3, 20)
Internal challenges for marketing, a European data point
Share of companies of the digital industry in Germany; n = 180, multiple answers; selected answers, labels translated from German
- Cost pressure54 %
- Automating marketing processes54 %
- Coordinating marketing processes37 %
- No AI strategy35 %
- Insufficient integration of AI34 %
Automation and AI are seen as the solution, but internally they are themselves part of the challenge.
Original: “Automatisierung und KI gelten als Lösung – sind intern aber selbst Teil der Herausforderung.”
2Trust is earned step by step
Gartner expects over 40 % of agentic AI projects to be cancelled by the end of 2027, because of escalating costs, unclear business value or inadequate risk controls. Many use cases sold as agentic do not need an agent at all, and its advice sorts the tools by the kind of work:(Gartner, agentic AI prediction 2025, press release)
They can start by using AI agents when decisions are needed, automation for routine workflows and assistants for simple retrieval.
State of Martech 2026 sees a similar order across all six categories of its survey: analytical AI is adopted first, generative AI second, and AI that acts on its own a distant third. Each step asks for more confidence. AI that reports what happened carries little risk; AI that drafts something for review needs editorial trust; AI that spends budget or speaks for the brand needs the trust of the institution.(State of Martech 2026, p. 33)
Where agents already act, the report finds a plain obstacle. Among the reasons teams hold back on agents in social media, the third is accountability: most organisations have not settled who owns the outcome when an agent gets it wrong, and without an owner nobody signs off.(State of Martech 2026, p. 45)
3What runs on each layer
Acontema organises its analysis in four levels, the same four as in its Analytics menu: Strategic, Operational, Performance and Business Outcome. On each of them some work runs on a schedule or on a trigger, and the result lands where the next level picks it up.
Automated work per layer
Strategic
- Competitor channels re-analysedDaily to monthly
- Search suggestions checked for content gapsNightly
- Trend discovery with sources and ideasOn a schedule
Operational
- A content plan writes titles and texts for a date rangeIn the background
- Drafts checked against campaign, audience and factsBefore review
- Approvals routed per language versionOn submission
Performance
- Monitoring runs; every article analysedDaily, every 3 days or weekly
- Topic trackers across owned, earned and paid mediaWeekly, monthly or quarterly
- Questions run against ChatGPT and GeminiWeekly or monthly
Business Outcome
- The analyst agent builds the analytics reportMonthly, quarterly or yearly
- Front-page reportOn schedule
- Any past period, with a cost previewOn demand
Watching agents, across all four
Collection, analysis coverage, brand accounts, content gaps, products, entities, chart data and print clippings. Mostly every morning; each with its own log.
Named people for every kind of task
A RACI role per person and a cadence: immediately, daily digest, weekly digest or inbox only.
What runs without being started by hand, per level, and what checks it. Cadences are set per project.
The point of the layering is the hand-over. A gap found at night on the strategic level becomes a content idea; an accepted idea becomes a draft in a content plan; the published piece is tracked in coverage and in AI answers; and the monthly report shows whether it moved anything. No step waits for someone to copy a result from one tool into the next.
4Agents that watch the work
Scheduled work fails quietly. A source changes its layout, a social account is renamed, an analysis stops halfway. Acontema runs watching agents over each project that look for exactly that and flag what needs attention:
- Collection. Every day: did the search for new articles actually run, and did it bring anything back? A quiet press and a broken search look the same otherwise. A weekly check looks for posts collected twice.
- Analysis coverage. Every day: does every social mention carry a current analysis? Missing ones are ordered by the agent itself.
- Accounts. Social accounts that carry the brand’s name but are not among its channels, such as a regional office or a dealer. The agent brings the list; a person decides which are the brand’s own.
- Content gaps, products and entities. New gap titles, new products and entity suggestions that a person should review.
- Chart data and print clippings. Charts that miss the dates they need, and uploaded files that do not look like print clippings.
Each agent shows its work: its schedule, its current verdict, its log and an estimate of the hours of work it saved in the last 30 days. Next to them the analyst agent takes the recurring tasks that produce something: the monthly, quarterly or yearly report and the front-page report, on schedule or on demand for any past period.
5Someone answers for every automated task
Automation without an owner is the failure the State of Martech authors describe. In Acontema every kind of automated task has named people, each with a RACI role (responsible, accountable, consulted, informed) and their own cadence: immediately, as a daily or weekly digest, or only in the inbox. What needs action lands in the project inbox under Needs action, with a due date.(State of Martech 2026, p. 45)
In a volatile market, accountability and clarity will be the cornerstones of competitive advantage for B2B leaders.
Cost is part of accountability too. Paid operations such as monitoring runs, platform searches, prompt tracking and scheduled reports show their price or a monthly estimate before they start, and operations that need credits do not start without them.
6From the surveys’ gaps to the platform
- Automating marketing processes
- Scheduled work on all four levels, from competitor re-analysis to the monthly report, with watching agents that check it ran.
- Rising resource needs for content and social media
- Content plans fill a date range with drafts in the background, written from the project context, translated and adapted per channel; comments under your posts are collected and classified.
- Coordinating marketing processes
- Responsibilities per post (RACI), approvals per language version, one inbox per project for what needs action.
- Processes not mature enough to scale AI
- Automated work is set up per level, not tool by tool: scheduled work, watching agents that check it ran, and named people for each kind of automated task.
- No AI strategy
- The four levels give the strategy its shape: what AI does on each, who owns it, and what it hands to the next level.
- Insufficient integration of AI
- AI runs inside the workflow, on one project context per brand or client, instead of in a separate chat window. Results go out as PowerPoint, PDF, Excel or through Zapier.
Sources
- The state of AI in 2026: On the road to ROIMcKinsey & Company (Dan Tinkoff, Lieven Van der Veken, Michael Chui, with Tara Balakrishnan), 25 August 2026. Online survey of 1,719 participants in 97 nations, 4 May to 8 June 2026; 36 % from organisations with more than $1 billion in annual revenue; weighted by each nation’s share of global GDP.
- Gartner 2026 CMO Spend Survey Finds CMOs Allocate 15.3% of Marketing Budgets to AI, But Only 30% Are Ready to Scale AI CapabilitiesGartner, Inc. (press release), 11 May 2026. 401 CMOs and other marketing leaders in North America, the United Kingdom and Europe, January to March 2026; the vast majority from companies with annual revenue over $1 billion.
- Marketing im digitalen Wandel: Zwischen Effizienz, Automatisierung & Wettbewerb. Lagebild der Marketingpraxis in deutschen Unternehmen 2026Bitkom e.V. (Bitkom Research), 12 February 2026. Online survey of 180 companies of the digital industry in Germany from the Bitkom network, calendar weeks 44 to 50 of 2025; in German, not representative.
- Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027Gartner, Inc. (press release), 25 June 2025.
- State of Martech 2026Scott Brinker and Frans Riemersma (chiefmartec, MartechTribe), May 2026. Survey of 208 marketing and marketing operations leaders on about 70 AI use cases, February 2026.
- Forrester’s 2026 B2B Marketing, Sales, And Product Predictions: B2B Companies Will Lose More Than $10 Billion Because Of Ungoverned Use Of Generative AIForrester Research (press release via Business Wire), 28 October 2025.
Quotes are verbatim; quotes in another language are shown in our English translation, with the original below. Page numbers refer to the printed page of the PDF. Statements about the platform describe what it does today.
See it on your own brand.
14-day free trial. No credit card required.