Vil du vide, hvilke AI-værktøjer der faktisk skaber målbar ROI for din virksomhed? Kenneth Dosanjh — AI-ekspert, revisor og forsker ved Erhvervsakademi København — hjælper danske SMV’er med at omsætte AI-eksperimenter til konkrete resultater på bundlinjen. Book en gratis konsultation i dag.
From Experiments to ROI:
How Kenneth Dosanjh Turns AI Into
Real Marketing Growth for SMEs
Most SMEs try AI, get excited, then quietly abandon it. AI expert Kenneth Dosanjh — accountant, researcher, and practitioner — explains exactly why that happens and how to fix it.
March 20, 2026
10 min read
Researcher, Copenhagen Business Academy
1. The Experiment Trap: Why SMEs Stall at “Interesting”
There is a pattern that AI expert Kenneth Dosanjh sees over and over in small and medium-sized businesses across Denmark: the experiment plateau. A business owner reads about generative AI, signs up for a trial, produces a few blog posts and some social media images — and then nothing changes. The tools sit idle, the subscription quietly renews, and AI remains a talking point rather than a business driver.
It is not a technology problem. The tools work. It is a strategy problem: most SMEs never build the bridge between “AI can do this” and “AI is generating measurable return for my business.”
Studies from Copenhagen Business Academy and similar institutions consistently show that SMEs which define success metrics before deploying AI tools are 3–4x more likely to report positive ROI within six months — compared to those who experiment without a measurement framework.
The difference between SMEs that get results and those that do not is not budget, technical skill, or the sophistication of the tools they choose. It is intentionality: defining what success looks like before the first prompt is written.
2. Who Is Kenneth Dosanjh?
Kenneth Dosanjh occupies a rare intersection: he is simultaneously a practicing chartered accountant, an AI expert advising SMEs on digital transformation, and an active researcher at Copenhagen Business Academy (Erhvervsakademi København) — one of Denmark’s leading business schools.
That combination matters. Most AI consultants either come from a pure technology background or a pure marketing background. Kenneth Dosanjh brings a third lens: the accountant’s demand for measurable, financially grounded outcomes. Every AI initiative he recommends must answer the same question that every good investment does: what is the return?
“AI without measurement is just cost. AI with the right metrics is investment. My job is to help SMEs make that shift — from spending money on tools to generating return on them.”
— Kenneth Dosanjh, AI Expert & Researcher, Revisorken
His research at Copenhagen Business Academy focuses on the practical adoption barriers that prevent small businesses from scaling AI beyond the pilot phase — and the organisational conditions that predict successful, sustained AI-driven growth. This academic grounding means his client work is not based on anecdote or vendor marketing; it is built on evidence.
Connect with Kenneth directly on LinkedIn to follow his ongoing research and practitioner insights.
3. The 3-Phase Framework: Experiment → Measure → Scale
The framework Kenneth Dosanjh uses with clients — and validates through his academic research — has three distinct phases. Most SMEs only complete the first one.
Phase 1: The Right Way to Experiment
Experimentation is not random play. Kenneth Dosanjh structures the experiment phase around a simple constraint: every test must connect to a specific task that currently costs the business time or money. Not “let’s try AI for blog posts” — but “we currently spend 4 hours and 2,000 DKK per month on blog production. Can AI cut that by 50%?”
This framing transforms experiments from entertainment into business cases. It also makes Phase 2 — measurement — straightforward, because the baseline is already defined.
Before any AI deployment, Kenneth Dosanjh asks clients to complete this sentence: “We currently spend [X hours / X DKK] per month doing [task]. If AI could reduce that by [Y%], our annual saving would be [Z DKK].” If you cannot fill in the blanks, the experiment is not ready to start.
Phase 2: Measurement That Finance Directors Respect
As an accountant, Kenneth Dosanjh insists on financial-grade measurement — the kind that would pass scrutiny in an annual report. That means tracking actual costs (tool subscriptions, human review time, prompt engineering hours) against actual outputs (content pieces, conversion rates, leads generated) and comparing both to the pre-AI baseline.
Phase 3: Scaling What Works
Most SMEs never reach Phase 3 because they try to scale everything simultaneously. Kenneth Dosanjh’s rule: only scale use cases that have demonstrated positive ROI in Phase 2. Pick the top one or two performers and build them into standard operating procedures — documented, repeatable, and measurable.
4. The ROI Metrics That Actually Matter
Not all AI metrics are equal. Kenneth Dosanjh divides them into three tiers based on how directly they connect to business value:
| Metric | What It Measures | Difficulty |
|---|---|---|
| Time saved per content piece | Hours reduced × hourly rate = DKK saved | Easy |
| Cost per lead (AI vs. pre-AI) | Total marketing spend ÷ leads generated | Easy |
| Content output volume | Posts/articles/emails produced per month | Easy |
| Organic traffic growth | SEO sessions 90 days post-AI deployment | Medium |
| Email open/click rates | AI-written vs. manually-written campaigns | Medium |
| Revenue attributed to AI content | Conversion tracking from AI-produced assets | Advanced |
Start with the “Easy” metrics in your first 90 days. They are immediately calculable and will give you the internal business case to justify further AI investment — whether to yourself, a partner, or a bank.
AI software subscriptions are fully deductible as operating costs in Denmark. Document your AI tools, their costs, and the business use cases they serve — Kenneth Dosanjh recommends treating AI as a line item in your monthly P&L from day one.
5. What Academic Research Reveals About AI Marketing ROI
As a researcher at Copenhagen Business Academy, Kenneth Dosanjh has access to emerging evidence on how Danish SMEs are actually performing with AI — not what vendors claim, but what the data shows. Several findings from his research and the broader academic literature are directly actionable for SME owners:
Speed to measurement predicts success
SMEs that establish baseline metrics within the first two weeks of AI adoption are significantly more likely to sustain use beyond three months. The act of measuring creates accountability and reinforces adoption.
Human-in-the-loop consistently outperforms full automation
Research consistently shows that AI + human review produces higher-quality outputs and better customer response than fully automated AI content — particularly for trust-sensitive industries like accounting, legal, and financial services.
Volume alone does not drive ROI
Producing more content faster does not automatically generate more leads or revenue. SMEs that see the strongest ROI focus AI on targeted content — specific search queries, specific customer segments, specific conversion points.
The biggest ROI often comes from operational AI, not creative AI
While AI-generated content gets the most attention, Kenneth Dosanjh’s research finds that SMEs often generate their highest ROI from operational AI: automated email flows, customer service responses, and CRM data processing — not blog posts.
6. The 5 Mistakes That Kill AI Marketing ROI
These are the five failure modes Kenneth Dosanjh sees most often — not hypothetical risks, but patterns observed repeatedly across Danish SMEs from 2024 to 2026.
Mistake 1: Choosing tools before defining use cases. Many SMEs subscribe to an AI platform and then figure out what to do with it. The reverse is correct: identify the three highest-cost or highest-friction marketing tasks first, then find the tools that address them.
Mistake 2: Skipping quality control. AI models produce confident-sounding errors. Every piece of AI-generated content — text, data summaries, email copy — requires human review before it reaches a customer. One factual error in a financial or professional context can undo months of brand building.
Mistake 3: Measuring activity, not outcomes. “We published 20 AI-generated posts this month” is an activity metric. “Our organic traffic increased 34% and generated 12 qualified leads” is an outcome metric. Kenneth Dosanjh insists clients track only the latter.
Mistake 4: Scaling too early. Doubling down on a tool before you have outcome data is how SMEs waste AI budgets. Three months of baseline data is the minimum before scaling any use case.
Mistake 5: Treating AI as a department, not a process. AI should be embedded in existing workflows — not siloed in a “digital team” or treated as a separate initiative. The businesses that get the best results are those that integrate AI at the workflow level: into their content calendar, their CRM, their email sequences.
7. Your 90-Day Action Plan
This is the exact 90-day plan Kenneth Dosanjh uses when onboarding a new SME client. It is built around the Experiment → Measure → Scale framework and designed to produce a defensible ROI number by the end of the quarter.
Days 1–7: Baseline Audit
Document the current cost (time × hourly rate + direct costs) of your top three marketing tasks. This is your pre-AI benchmark. Without it, you cannot calculate ROI.
Days 8–30: Controlled Experiment
Deploy one AI tool against one specific task. Run it in parallel with your existing process for four weeks. Track output quality, time spent, and cost per unit.
Days 31–60: Measure and Decide
Compare AI results to your baseline. Calculate the ROI number. If positive, continue. If negative or neutral, change the tool, the use case, or the prompting strategy — and re-test for 30 more days before concluding.
Days 61–90: Scale One Use Case
Take the single highest-ROI use case and build it into a documented, repeatable workflow. Add a second AI tool only if the first has positive results. Add outcome metrics to your monthly business review.
8. Conclusion
The gap between SMEs that get real value from AI and those that don’t is not about technology, budget, or sophistication. It is about the discipline to measure. The businesses that treat AI adoption the way a good accountant treats any investment — with clear baselines, defined success metrics, and honest evaluation — are the ones that turn experiments into growth.
Kenneth Dosanjh’s dual role as a practising accountant and academic researcher gives him a perspective that is unusual in the AI advisory space: he holds every tool recommendation to a financial standard. If the numbers don’t work, the tool doesn’t stay. If they do, it gets built into the business permanently.
The question for any SME owner is not “should we use AI?” It is: “which specific task, measured how, with what success threshold?” Answer that before you open a single new AI subscription — and you will be ahead of 80% of your competitors before you write your first prompt.
Follow Kenneth Dosanjh’s ongoing research and practitioner insights on LinkedIn, or book a free consultation to discuss your SME’s AI strategy directly.
Ready to turn your AI experiments into ROI?
Book a free 30-minute consultation with Kenneth Dosanjh to build your personalised AI measurement framework — and find out exactly which tools deserve a place in your budget.
