Last updated 21 September 2026 · 14 min read · By Mattias Loxi Co-founder, Cinode
Skills Management is the practice of systematically mapping, developing and deploying the skills in an organisation. For a consulting firm it comes down to three things: knowing what skills you have, knowing what the market is asking for, and knowing what the gap between them costs you.
This is not an HR topic. It is the business model. You sell expertise by the hour. If you don’t know precisely what you have to sell, you sell it less well.
A working process answers four questions:
| Question | What it governs |
| What skills do we have? | What you can sell today, and who can staff what. |
| What skills are in demand? | Which assignments and clients to pursue. |
| Where is the gap? | What to train, hire or buy in. |
| How are skills developing? | Utilization, rates and attrition over time. |
Five things worth knowing
Skills matrix — an overview showing which skills each person has, and at what level.
Skill set (also requirement profile) — a group of skills with required levels, describing what a role or assignment demands.
Skills gap analysis — the comparison between the skills you have and the skills you need.
Utilization — the share of a consultant’s available time that is booked on client work.
The bench — consultants who are available but not booked on an assignment.
Win rate — the share of bids or assignment requests that convert.
Subcontractor — a consultant from another firm you bring in to deliver under your name.
A technical skill used to hold its value for eight to twelve years. By 2030 that drops to somewhere between two and five, according to Gartner’s Predicts 2026: AI’s Impact on the Future of Workforce.
But not all skills age at the same rate. The distinction between durable and perishable skills — a framework from Chief Learning Officer — puts tool-specific abilities at a half-life of under two and a half years, while durable skills like judgement and communication hold for over seven and a half. And perishable is precisely the category your CVs and consultant profiles lean on hardest, because a named tool is easy to write down and easy for a client to search for.
How large is the shift overall? Roughly 39 percent of today’s skills will be transformed or outdated within five years, according to the World Economic Forum’s Future of Jobs Report 2025. In AI-adjacent fields it moves faster still.
The pace is actually stabilising
The share of skills expected to change by 2030 has fallen — from 44 percent in 2023 to 39 percent in 2025. The reason is that more organisations are doing something about it: half the workforce now undertakes training as part of a long-term strategy, up from 41 percent two years earlier.
So the conclusion is not that everything is collapsing. It is that firms with a structure for continuous skills development are keeping pace, while those without one fall behind faster and faster.
Cinode publishes two recurring studies of the Nordic consulting market: a quarterly analysis of listed Nordic consulting companies, which still shows no real turnaround, and an annual review of more than 300 privately held Swedish consulting firms, based on published annual accounts and run with the same method since 2014.
The 2025 review shows an average operating margin of 4.4 percent, down from 5.7 the year before and 7.6 three years ago. A quarter of the firms reported break-even or a loss. One in five lost more than five percentage points of margin.
At that level a bad hire costs more than it did three years ago, and a skill you have but can’t find gets more expensive with every request that goes to someone else. Skills data isn’t an HR investment at those margins. It goes straight to the result.
Without current skills data you cannot work preventively. The firm reacts to what happens instead of steering it.
Do consulting firms have a capacity problem? Not really. Do they have a visibility problem? Often. Utilization measures hours, not what those hours can deliver — so people sit on the bench while the firm turns down work. What follows from that?
None of these has its own line in the income statement. That’s why they rarely get addressed: the cost is booked as something else. How to put numbers on it is covered in our guide to utilization in consulting firms.
The process works whether you have 20 consultants or 500. Only the pace differs.
| Step | What it gives you | |
| 1 | Skills inventory. | You know what you have to sell. |
| 2 | Your skill sets. | You know what you need. |
| 3 | Gap analysis. | You know where you’re exposed. |
| 4 | Development plans. | Closes the gap internally. |
| 5 | Hiring and partners. | Closes the gap externally. |
A structured account of the skills that exist in the firm, per individual. This is the foundation for everything else — without it, the rest is guesswork.
What to include:
That last point is almost always forgotten, and it is the most valuable. Skills you have but nobody wants to use are hard to keep selling.
For consulting firms with technology vendors as partners, certifications tie directly to revenue. Miss a renewal and you can lose partner status — and with it both your rate level and your eligibility to bid.
Telia Cygate, one of Sweden’s leading systems integrators with around 650 employees across 20 offices, keeps full visibility of which certifications exist, when they expire and which need renewing. They act ahead of time instead of reacting once it’s already too late.
The barrier is effort, not intent — nobody fills in a blank template voluntarily. Start from material that already exists: import existing CVs or LinkedIn profiles and let the consultant complete them. Suggest skills from a shared taxonomy rather than offering a free-text field, or you’ll end up with five spellings of the same technology.
But what really decides the outcome is whether the data gets used. Softhouse describes the mechanism well: once the data is available to everyone, a virtuous cycle takes hold where quality improves and more people add more.
That is the heart of the adoption problem. Skills data that is collected but only read by management dies within a quarter. Data the consultant benefits from directly — better assignments, internal visibility, a CV generated automatically — maintains itself.
Profiles are built by importing existing CVs or LinkedIn data, so consultants complete rather than start from scratch. Skills are suggested from a shared taxonomy, which prevents the same technology being entered five different ways. Certifications are logged with expiry dates and reminders. CVs and consultant profiles are generated automatically from the data — so the consultant who keeps their profile current gets a finished CV back.
Groups of skills describing what a role, an assignment or a client type actually requires. The inventory shows what you have; the skill set shows what you need. The gap between them is the whole point.
Start from market demand, not from your existing skills.
Then define the required skills, with levels. A skill set for a front-end developer at your firm might include JavaScript, React, CSS, test automation and agile methods — but which levels are required depends on your assignments, not on a generic template.
Tip: Build your first skill sets from your ten most recent client requests. That produces a profile reflecting real demand rather than internal assumption.
How this works in Cinode
Skill sets are reusable and tied to roles or assignment types. The same profile can then be used to match internally, to write a job advert and to answer a tender.
Here the inventory is set against the skill sets. The result shows where you’re strong, where you’re exposed, and where you hold skills nobody is asking for. This is the step where data becomes a basis for decisions.
Five questions to answer:
That last question gets overlooked most often, and it matters most. A single AWS architect is not a skill. It is a risk.
Esri Sweden, the country’s leading GIS software provider, works exactly this way: identifying skills gaps before they become urgent and matching the right person to the right assignment. That is where the long-term effect comes from.
One more thing. Don’t only compare against your own skill sets — compare against the market. Which skills are appearing in client requests that weren’t there a year ago? That is the earliest signal you get of a technology shift, earlier than industry reports and earlier than competitors manage to adapt.
The skill set is set against actual skills data and the gap is shown directly, for one person, a team or the whole firm.
Search matches on meaning, not on wording. Enter “cloud architect with financial services experience” and the consultants who fit are surfaced, even if they described their skills differently. The prerequisite is that there is something in the profiles to match against — good search doesn’t compensate for thin data.
Individual plans connecting the consultant’s development to the firm’s skills needs.
This is where most initiatives fall down. The gap analysis gets done, it looks good, and then nothing happens because everyone is booked.
What actually works:
Start with the individual, not with the gap. Ask what the consultant wants to develop, then match that against the firm’s needs. Forced development produces worse results than none.
Vincit built its entire matching process around that principle. Nobody should be assigned to a project at random. The assignment should move personal development forward. Alongside skills, they therefore make three things visible: employees’ personal goals, their development plans, and the type of project they want to work on.
This has an effect in two directions. The firm matches better and the employee experience improves, which in turn shows up in attrition.
Use assignments as development. Prioritise new, stretching work over long assignments where the consultant coasts. A comfortable three-year posting is the most common reason a good consultant leaves.
Let specialists teach. Who is at level 2 aiming for 3? Who on the team can coach them there? Internal knowledge transfer is cheaper than courses and builds culture as a side effect.
Set a deadline and review quarterly. A development plan without a date is a wish list. Follow up with concrete goals, give feedback, and set new goals for the next period.
Rotate assignments. Don’t leave the same consultant with the same client doing the same tasks indefinitely. Grow someone else into it.
Make a point of new skills. Recognise the people who gain new or deeper expertise. Communicate it internally and externally. It signals to both employees and the market that you mean it.
Personal goals, preferred roles and development plans sit on the same profile as the skills. Whoever staffs an assignment sees not just who can, but who wants to.
Link the follow-up to performance reviews and the plans get reviewed four times a year instead of written once and forgotten.
The plan for the gaps you can’t close internally.
Once the gap is known, hiring becomes a planned activity rather than a panic measure. The difference in both cost and quality is substantial.
Hire proactively. Build a pipeline of candidates in your specialist areas and stay in contact. Map their skills before the need arises — then staffing takes days rather than months.
Use subcontractors strategically. Sooner or later a deal arrives with skill requirements you can’t deliver in time. At that point you can neither hire nor train. An established partner network is the difference between taking the deal and declining it.
Treat subcontractors as an asset, not a threat to the brand. The fastest-growing firms combine employees, partners and freelancers.
Avoid person dependency here too. Whoever runs the partner network should work in a structure someone else can take over.
Forefront Accelerate, with around 150 consultants on assignment, left the spreadsheet behind so they could reach their partner network with relevant requests. They gained more subcontractors and partners — and considerably more relevant responses, which produced more deals.
Fujitsu in Sweden sees the same mechanism from the other side. Their clients often ask for niche roles that neither Fujitsu nor their regular partners hold. Through the network they reach specialist firms with unusual expertise, and it has helped them win more business.
The partner network sits in the same system as your own skills data, so a request can be matched against employees and subcontractors at once. Assignment requests can be sent to the whole network and answered with finished consultant profiles. Requests arriving by email are matched automatically against available consultants, which shortens response time — the metric that decides most deals.
With the steps covered, you can place your own firm. Most consulting firms recognise themselves in one of five stages, and which one you’re at determines what to do first.
| Stage | Characteristics | What happens as you grow |
| 1. Person-dependent | The skills overview lives in one or two people’s heads. | Breaks at 30–50 consultants, or when that person leaves. |
| 2. Documented | CVs in Word, skills list in a spreadsheet, updated sporadically. | Updates never catch up with reality. |
| 3. Structured | Skills data in a system, searchable, reasonably current. | Works, but only reactively when a request comes in |
| 4. Directive | Skills data drives hiring, training and which assignments you pursue. | Scales. |
| 5. Predictive | You see the gap before it appears, and match automatically against incoming requests. | Competitive advantage. |
The move from stage 2 to stage 3 is the one most firms have ahead of them. Omegapoint, a fast-growing cybersecurity specialist, is an example of exactly that transition — they left the spreadsheet behind when it stopped keeping up.
Don’t skip stages. A firm at stage 1 that buys a system without changing how it works lands at stage 2 with a higher cost base.
What comes next depends on where you are. At stage 1 or 2: finish the skills inventory before building anything on top of it — everything else rests on it. At stage 3 or 4: the data already exists, and the work lies in using it preventively rather than at every incoming request.
1. Starting with the tool. A system doesn’t solve a structural problem. Define what you need to know and why, then how you collect it.
2. Over-detailed taxonomy. Fifty skills with seven levels each produces data nobody fills in. Start coarse and refine where it matters.
3. Self-assessment only. Useful for finding interest and direction, unreliable as a basis for selling. Validate the skills you actually sell on.
4. Forgetting the subcontractors. If a third of your delivery capacity is external and you only track employees, you’re tracking two thirds of the business.
5. No owner. Skills data that is everyone’s responsibility is nobody’s. Assign it to someone on the leadership team.
6. Treating it as a one-off. Without a rhythm — quarterly updates, an annual review of skill sets — the work is wasted within a year.
Skills management should be visible in the numbers. Track these:
| Metric | What it tells you |
| Utilization | Whether you’re matching skills to demand. The best in the industry stayed above 90 % in 2025. |
| Share of profiles updated this quarter. | Whether the data is decision-grade. |
| Response time to a client request. | The most direct effect of searchable skills data. |
| Win rate on bids. | Whether you’re matching the right consultant to the right request. |
| Business-critical skills held by one person. | Your exposure. |
| Attrition. | Whether the development plans are working. |
| Share of assignments staffed internally vs externally. | Whether the gap analysis is accurate. |
Kvadrat, Sweden’s largest network of independent consultants, points to the metric that moves first: when skills, CVs and sales activity sit in the same solution, they can respond to requests faster and match the right consultant to the right assignment.
What’s the difference between skills management and skills development? Skills development is the activity — courses, certifications, mentoring. Skills management is the system around it: knowing what skills exist, what’s needed, and which development efforts are actually the right ones.
How often should skills data be updated? Skills and levels at least quarterly, ideally alongside performance reviews. Certifications continuously, with expiry monitoring. Skill sets annually, against what client requests actually contain.
How many skill levels should we use? Three to five. Fewer than three says nothing about depth; more than five makes assessment unreliable because nobody can distinguish level 6 from level 7. Four is a solid starting point: knows the basics, works independently, leads others, is an expert.
Does self-assessment work? For capturing interest, breadth and direction: yes. For guaranteeing a consultant can handle an assignment: no. Validate the skills you actually sell on.
What does it cost not to know? Count three items: avoidable bench time, deals lost to a faster competitor, and hires that could have been internal training. In a firm of 100 consultants running a 4.4 percent operating margin, a few weeks of unnecessary bench time is enough to consume the year’s result.
Do small consulting firms need skills management? Under 20 consultants you can manage with simple tools and close dialogue. The point is to have the structure in place before you grow — the transition is considerably harder at 60 consultants than at 20.
How do we get consultants to actually fill in their skills? Lower the barrier, show the benefit, make it visible. Import existing CVs instead of handing over a blank template. Make sure the consultant gets assignments matching what they want to work on. Highlight expertise internally and externally. Compulsion works once.
Does this work in large organisations? Yes, and the need grows with size. HiQ has 1,700 experts across Sweden, Finland and Germany. Consid has more than 2,000 employees. At that scale the skills overview can no longer live in anyone’s head.
Just starting out? Begin with one team or one skill area, not the whole firm. It’s cheaper to learn where the resistance sits with twenty people than with two hundred.
Have the data but not using it? Work backwards. Take your ten most recent client requests and try to answer them with the data you hold. What you can’t answer is your real requirement specification.
