Fractional Chief Innovation Officer

Corporate innovation and venture portfolios, run on evidence rather than sponsorship.

Innovation as a capability everyone can afford, not only the organisations with a department for it.

25+ years building brands, technology teams, and ventures. Outside of running an AI-native product development factory and an AI-native venture factory, Mark takes on a small number of fractional engagements with corporates, funds and accelerators: helping leaders decide what to build, what to stop, and what their data has to be ready for first.

Portrait of Mark Morris
25+ years building South Africa, UK & USA Founders, funds and listed enterprises Co-founder, FiftyKnots & AI Acumen Cape Town, ZAAvailable: Selective

Measured, Not Estimated

Consultants tell you what to do. Operators show you how. Builders install machines that keep the engine running after they leave.

The ones worth hiring have already done it to their own business. He did it to two: both ran the traditional way, and he rebuilt them AI-native himself, as a business user rather than a sponsor, for roughly a third of what a conventional team was quoted. The numbers below come from the portfolio that came out of it, and they include the ventures that were stopped.

25+ Ventures built and governed
50%+ Closed, not quietly parked
3 Exited by sale or merger
22 Validated through the platform
7 weeks Signed NDA to shipped MVP

Twenty-five years of ventures, built and governed. Those not closed or exited sit at known stages: scaling, product-market fit, launch, validation, seed. The seven weeks is the founders’ own count, from the date on the NDA.

Both platforms are live and processing founders now, at ai.acumen.zone and app.fiftyknots.com. The rebuild took longer on the calendar than the conventional estimate, and cost roughly a third as much. That is the honest trade.

Data first. Then strategy. Then the systems.

Mark has spent 25+ years moving between brand and consumer strategy, the technology teams that turn ideas into working products, and, more recently, a portfolio of ventures built with AI from the ground up. He builds forward, not backward: data first, strategy second, systems third. Most organisations do it in reverse, a strategy deck first, then a product, then a hunt for data to justify both. It's slower for the first month and faster for every month after.

That order-of-operations shows up in two studios he co-founded. FiftyKnots is a venture studio that takes founders from a raw idea to a launched, evidence-tested business in weeks rather than years. AI Acumen works the same discipline in the other direction, helping companies in growth, scale or transformation turn scattered data into governed, AI-ready foundations before layering products and workflows on top. Both are proof the method works, not the offer itself: a fractional engagement buys Mark's time and judgment. If a conversation later turns into a build, that's a separate decision with the right team for it.

His fractional work sits at board and executive level, with two kinds of organisation: South African mid-market and enterprise companies with a real mandate to act on innovation and AI, and the funds and accelerators governing portfolios of early-stage ventures. In both cases a mandate to act, not a request for a workshop. He has designed innovation for companies like FiberLean, Altron, Mr Price and KPMG, across technology and SaaS, retail, manufacturing and professional services.

Ventures

An AI-first product development studio. Its own inbound sales desk runs in production on the platform it sells, so the studio is its own first customer.

A venture studio taking founders from a raw idea to a launched, evidence-tested business in weeks rather than years. Twenty-two ventures validated and supported so far.

Escrow-based marketplace for home improvement work. NDA to shipped MVP in seven weeks and one day; launched August 2026 and trading.

Runs governance for corporate innovation portfolios: the same discipline he brings into a fractional seat, productised for the companies he sits with.

The Orchestrator

One triage engine applied three ways: inbound sales, investment intake and publication. It classifies what arrives, scores it against fit, and routes it to the right next step.

Co-founder and non-executive director. Automates growth with AI strategy, centralised control and scalable execution for marketing teams.

GigTribe

A platform assembling verified, high-performing remote teams for global projects, cutting lead times and raising execution confidence.

Co-founded a smartphone-first workforce time and attendance platform, replacing expensive hardware with mobile tools for SMEs.

Engagement Models

Six ways to work together.

Sized to how deep the problem needs to go, from standing advisory to a fixed-scope diagnostic.

01

Fractional Chief Innovation Officer

The question this seat answers: "How do you build a data- and AI-driven innovation portfolio?"

Owns the innovation agenda at exec level: classifying the portfolio honestly, an assumption registry per initiative so gate decisions rest on evidence rather than sponsorship, the AI and data-readiness agenda, build-vs-buy calls, an honest kill rate, and board reporting that shows commercial outcomes rather than activity.

1–2 DAYS/WEEK RETAINER
02

Advisory

Standing access for a CEO, CIO or founder who wants an operator on call: a sounding board on AI decisions, product direction, and venture calls. Quiet, honest, no deck required.

A FEW HRS/MONTH LIGHT RETAINER
03

AI & Data Readiness Diagnostic

An evidence-graded answer to "is our data ready for what we want AI to do?", with a sequenced plan: foundation gaps, governance, quick wins, and what not to build yet.

2–4 WEEKS FIXED SCOPE
04

Innovation-Ready Data Foundation

The question it answers: "Is your data and AI foundation innovation-ready?"

Most innovation portfolios fail on context, not ideas. This installs a provenance map, a governed context layer over raw sources, lifecycle rules, and decision records that make every kill-or-cultivate call auditable.

3–6 WEEKS FIXED SCOPE
05

Venture/Product Evidence Review

An operator's read on a venture, product or portfolio before money moves: what the evidence supports, what it doesn't, and the honest kill-or-cultivate call.

FIXED SCOPE
06

Custom Playbooks

Scoped work on corporate innovation or AI-assimilation, built to the specific shape of the problem rather than a standard package.

SCOPED ON REQUEST

The Method

Most innovation portfolios are a list, not a system.

A list gets reviewed. A system makes decisions. The work is turning one into the other, and it runs in the same order every time.

01

Classify the portfolio honestly

Every initiative gets named for what it actually is: a core improvement, an adjacent bet, or a genuine new venture. Most portfolios have never done this, which is why they get governed as though everything in them carries the same kind of risk. It does not, and the same gate cannot serve all three.

02

Write down what has to be true

Each initiative gets an assumption registry: the things that must hold for it to work, owned by a named person, dated. Without it, a gate review is a discussion about confidence. With it, a gate review is a check against something written down before anyone knew the answer.

03

Grade the evidence, then gate on it

Each initiative faces a gate with a defined evidence bar. Pass, hold, or stop. The bar is set in advance, and it is the same bar for the sponsored project and the unsponsored one. That is the entire mechanism, and it is the part organisations skip.

04

Make the stop decisions hold

Anyone can run a portfolio. The test is whether initiatives with a sponsor ever get stopped, and whether the person in the seat has the standing to make that call stick a month later when the sponsor comes back. Stopping is the discipline the rest of it exists to serve.

05

Report commercial outcomes, not activity

Board reporting that says what was decided, what it cost, what was learned and what stopped. Not a status deck. The output of governance is decisions, so that is what gets reported.

None of this is theory. It is the discipline I run my own venture portfolio on, and it is why more than half of the ventures on that list were closed.

“I'd rather tell you the truth than send an invoice for optimism.
AI made the inference cheap. It did not make judgment, foundations, or expertise free.

How It Works

From first conversation to embedded work.

01

Conversation

An initial, unstructured discussion: what's stalled, what's been tried, and whether it's worth scoping anything at all.

02

Diagnostic

A short assessment of innovation readiness and data/AI foundations, using a Lean Canvas and Strategyzer-based framework rather than assumption.

03

Structure agreed

Retainer, advisory, or a fixed-scope project: whichever model fits the problem, with time commitment and cadence set upfront.

04

Embedded work

Evidence-based execution: validated decisions before capital or headcount is committed to them.

Results

A portfolio that grows into its own DNA.

Innovation theatre is easy. Evidencing success is what really matters.

01

Signed NDA to shipped MVP in seven weeks and one day

Nathan Filen and Anopa Chipangura, co-founders of Khaya Crew
Nathan Filen and Anopa Chipangura, co-founders, Khaya Crew

Khaya Crew is an escrow-based marketplace for home improvement work in South Africa. It went from signed NDA to shipped MVP in seven weeks and one day.

The founders did that arithmetic themselves, unprompted, counting from the date on the NDA. It is their number, not ours, which is the only kind worth quoting.

It launched on 13 August 2026 and is trading.

02

25+ ventures, and more than half were closed

Most innovation track records show you the winners. This one includes the ventures that were stopped, because the stopping is the discipline.

Across 25+ ventures built and governed over as many years, 3 exited by sale or merger and more than half were closed. The rest sit at known stages: scaling, product-market fit, launch, validation, seed.

Anyone can run an innovation portfolio. The test is whether initiatives with sponsors ever get stopped, and whether the person in the seat has the standing to make that call hold. Every closed venture on that list was a decision someone had to make and defend.

03

Fifteen years inside one enterprise

FiberLean Technologies is a UK nanotechnology business, spun out of a joint venture between Imerys and Omya. The relationship started in 2009 and is still running.

The work covered a full digital transformation: Microsoft Business Central as the finance and operations foundation, a cloud-only architecture at a time when that was still an unusual decision for a manufacturer, and specialised line-of-business applications layered on top.

"The only one of nine ERP implementations they'd been involved in that finished on time and within budget."

FiberLean's Chief Financial Officer put that on record in a 2018 letter of reference, quoting the executive who oversaw the programme. The architecture held: FiberLean didn't hire its first internal IT programme manager until five years after go-live.

04

The system he sells is the system he runs

AI Acumen is an AI-first product development studio, helping companies in growth, scale or transformation turn scattered data into governed, AI-ready foundations before layering products on top.

Its own inbound desk runs in production on the platform it sells. Enquiries arrive, get scored, and get routed by the same machine a client would buy. The studio is its own first customer, which is the one product test that cannot be staged.

That same engine now runs the desk behind FiftyKnots and IPnition as well. One system, three businesses, none of them a demo.

Live at ai.acumen.zone, fiftyknots.com and ipnition.com.

05

Twenty-two ventures through one machine

FiftyKnots is a venture studio that takes founders from a raw idea to a launched, evidence-tested business in weeks rather than years.

Twenty-two ventures have been validated and supported through its platform. The kill decisions in this portfolio were made with it, not around it.

Live at app.fiftyknots.com. Both studios are Mark's own companies, so read these two as working demonstrations rather than independent references. The independent ones are above.

06

What it costs to learn this

Skills Cafe went through three documented false starts. It launched as a contracting platform, pivoted to recruitment, then changed direction again without restarting validation from the beginning. The lesson from the third mistake was the expensive one: when you change what you're building, you have to redo the work that justified it.

That's when the approach changed to design-led, and it hasn't changed back. Mark didn't read this in a book. He paid for it, and now brings it into rooms where the same mistake is about to cost considerably more.

In Their Words

Where the work happens

Strategies come alive when they leave the designers' minds, and run as systems in client hands.

The fractional work is a deliberate allocation of focus outside the studios: different industries, different constraints, different evidence. It runs both ways: the corporates Mark sits with are the same organisations IPnition was built for, and the funds are governing the same kind of portfolio FiftyKnots runs.

FAQ

Before you reach out.

No. This is Mark, in a personal capacity. He co-founded FiftyKnots (a venture studio) and AI Acumen (an AI-native product studio), and they're the proof the method works. A fractional engagement buys his time and judgment. If it later turns into something that needs building, that's a separate decision and a separate conversation.
Having data isn't the same as being AI-ready. AI needs governed context: trusted definitions, quality, lineage, semantic models, secure access. Without that you get impressive demos and unreliable decisions. The diagnostic exists to answer this honestly before you spend on the exciting part.
You can, and most do, and it shows. Strategy derived from evidence beats strategy imposed on assumptions. Data first, then strategy, then systems. It's slower for the first month and faster for every month after.
Both exist. Mark has measured numbers from his own build: a build costed at six people for six months and $450,000, delivered instead by one and a half people in under nine months for roughly a third of that, on a platform now past its two hundred and fiftieth feature. Under $10,000 of it was AI inference and about $3,000 a month runs the services; the rest is his own time. AI made the inference cheap. It didn't make anything else free. He'll tell you which parts apply to your business and which don't.
That an innovation officer's job is to generate ideas. Ideas are the cheap part. The job is governance: deciding, on evidence, which initiatives earn the next rand and which stop, and having the standing to make the stop decision hold.
Whatever the engagement model specifies, agreed upfront rather than negotiated after the fact. A fractional retainer runs 1–2 days a week. Availability is kept deliberately limited so that number holds, not stretched thinner to accommodate one more logo.
No. It needs to be governed and trustworthy for the specific decision or use case in front of you, not enterprise-wide and pristine. Waiting for a perfect data estate before starting is usually just a slower way to fail. The diagnostic exists to identify which slice actually needs the work first.
That gets named explicitly when the engagement structure is agreed, not discovered after something breaks. A recommendation without a named owner isn't governance, it's a slide. Part of the job is making sure that question has an answer before it's needed.
The board has the final call, always. The job is to make sure that call is made on evidence rather than sponsorship or sunk cost, and documented well enough that whichever way it goes, it's defensible. Disagreement is fine. An undocumented decision is the actual risk.

Start a conversation.

Send an outline of the problem you're looking at, and Mark will come back with next steps directly.

Typically a response within two business days.

Your name, email, company and message are used only to answer this enquiry. They are sent to Mark's own mailbox through Resend, the mail provider for this site, kept until the enquiry is closed, and are never sold, shared, or added to a mailing list. Write to mark@markdmorris.com at any time to see what is held or ask for it to be deleted.