CIO explains: the decisive factors that set top data partners apart

🇫🇮 This blog is also available in Finnish (original version); please find it using the language switcher on the top right corner.

What do CxOs really value in a data partner?

More and more companies see data as a strategic asset, but far fewer know how to use it consistently. We interviewed Pete Nieminen, member of Recordly’s board, who has a deep understanding of the realities of buyers, strategy partners, and the technology industry. The result was an honest, hands-on perspective on where data companies succeed, and where they still stumble. This interview provides you with insights on the role of a data partner, the limits and challenges of data consulting, and how a C-level buyer can assess a data partner’s reliability and fit.

The blog in a nutshell 🌰

• Data partners must go beyond tech fixes and focus on creating long-term business value by aligning with the customer’s growth goals and challenges.

• CxOs value partners who challenge thinking, not just execute. The best data partners assess the true root of a problem and tailor the solution collaboratively.

• Strategic data use is often too narrow. Many companies silo data, focus only on reporting, or rely on outdated or irrelevant sources, especially in the GenAI era.

• Trust is built on reputation, transparency, and keeping promises. Buyers listen to peer networks, value humility, and expect accuracy in time and budget.

• Mature AI solutions are those with controllable risk. Companies should maintain critical thinking and ensure data quality before deploying GenAI solutions.

Can you briefly tell us who you are and about your current role on Recordly’s board?

When discussions with Recordly began, my role was defined as an advisor from the perspective of IT leadership and the buyer. My background is extensive in service development, productization, and sales, and in my board work I’ve mainly helped clarify what services should look like and how Recordly’s expertise should be communicated so that the customer clearly sees its value.

What first intrigued me about Recordly was the people, their expertise and deep technological understanding. On the board, I bring a strong CxO perspective: how leaders interpret things and what conclusions they draw.

I also enjoy contributing to the practical execution of strategy. I’m not just attending meetings to look at numbers, because I want to help solve real challenges and improve operations.

A modern board must be able to support a company in all situations and act as an active partner in development work. Networks play an important role in Finland and internationally; the best collaborations and projects originate from strong networks.

How can data companies create long-term value for customers instead of just solving technical problems?

Creating value requires more than just a technology partnership. The biggest challenge for data companies and consultancies is customers' widely varying level of data maturity. The differences can be huge: in some organizations, data is a strategic asset on the management agenda, while in others it’s just a tool for a small group of specialists. In these cases, the data partner must be able to operate at different maturity levels while also helping to raise the customer’s internal capabilities to a new level.

Today, every management team, CEO, and board should understand that data is a central factor in nearly all business development. Data consultancies like Recordly have deep technical expertise and an exceptionally strong understanding of how data benefits the business. The challenge – and the opportunity – is to make the message clear so it resonates even with leaders who are unfamiliar with data topics.

Long-term value is created when a data partner understands the customer’s growth goals and business challenges and can clearly show how data supports those goals and improves operations. I also value that Recordly has strong brand expertise: many consulting companies advertise brand agnosticity, but from a buyer’s perspective this can mean you need several partners – and then you risk inconsistent expertise and clumsy collaboration. Brand dependency can actually be a strength: the ability to combine knowledge of data and AI with practical technologies and recommend the best solutions without pushing them. Concrete examples and clear reasoning build trust.

Reputation is built based on actions. This requires courage, even in difficult situations. If a customer wants to move forward with a solution that differs from the consultant’s recommendation and the consultant believes it carries a significant risk of failure, it can be both ethically and commercially justified to decline the project. Long-term success is built on honesty and the ability to make hard decisions – even if that means turning an opportunity down.

What are the most common gaps companies experience when trying to use data strategically?

A typical blind spot is seeing data use in strategy too narrowly – only as reporting and analytics. Often the focus is on developing better reports and deeper analysis, while the data itself remains unchanged. There’s a lack of understanding that the value of data comes from cleaning, enriching, and broadening its use. 

For example, a company that uses only financial data in its strategy is limiting itself. Nowadays we know that customer, production, ESG, and people & culture data can add significant value to decision-making. If a strategy is built on one single data source, the outlook will inevitably be confined. Management should ask for better data rather than better reports.

For long-established companies that have collected massive amounts of data, the GenAI hype adds complexity: much of the data may be useless, incorrect, or simply outdated. For example, in market data, only the last 18 months may be relevant and anything older is essentially ancient history.

The most common pitfall is data silos. In many organizations, each specialist is responsible only for their own domain’s data, and there’s no role or structure to manage and integrate the organization’s entire data landscape. As a result, different units draw conclusions from their own data, but no one combines it into a full picture or refines it for strategic needs. Strategic data comes from combining data sets and analyzing them in new ways.

Another common challenge is roles and responsibility. Many organizations have appointed a Chief Data Officer or similar, but history shows that some of these titles can be short-lived – as happened with many CDO roles while the function itself lives on and merges into the business. It’s important to ensure that responsibility for strategic data use is not just formal, but has genuine resources and influence.

A third pitfall is disconnecting business understanding from data work. If data teams don’t understand the business’s goals and context, they may produce technically impressive but commercially irrelevant analysis. For long-term value, it’s critical that data specialists and business leadership work closely together and speak the same language.

What do C-level leaders look for when evaluating external data partners?

C-level leaders look for a data partner’s ability to solve real problems and challenge the customer’s thinking when needed.

Typically, a data partner is sought for two reasons:

1. The business needs data expertise it doesn’t have enough of internally.

2. Solutions are needed where business understanding and technical data expertise are combined, but internal capabilities don’t cover both.

The process often starts with the customer defining the problem. A good partner doesn’t just start solving it right away – they assess whether it’s defined correctly and what’s the best way forward. What’s valued is a clear vision and a concrete way to proceed, but not a rigid, one-size-fits-all package. Especially in data use, customer needs vary so much that the best solutions are always defined and tailored together. At the same time, it’s good to see proof that similar solutions have been implemented before.

What annoys customers?

Many consultants suggest starting with an assessment. But this can trigger resistance if it means multiple consultants doing the same interviews with the same stakeholders and collecting the same information repeatedly. From the customer’s point of view, it’s more efficient if one consultant gets familiar with the business and its challenges, and any further assessments are done internally within the consultancy – without burdening the customer with repeated meetings.

A good data partner knows how to combine strategic thinking with technical expertise, and respects the customer’s time. When these come together, the partner becomes a trusted advisor, not just a project supplier.

What makes a data partner trusted from a leader’s or buyer’s perspective?

Trust comes from reputation, transparency, and keeping promises.

First, reputation. C-level leaders get recommendations primarily from peers, vendors, and their own networks. Word spreads quickly about who’s worth calling – and who’s not. A good reputation is built on how well a partner has delivered real projects.

Second, transparency and humility. A trusted partner doesn’t pretend to know things they don’t. They’re willing to ask, listen, and learn from the customer. Even top experts can lose their credibility if they give the impression that they know everything. Mutual respect is essential from the start.

Third, keeping promises – especially on timelines and budgets. For example, there’s a Finnish company known for never exceeding its T&M (time-and-materials) estimate; it’s their principle. Conversely, I’ve seen cases where the budget exploded in the early stages with the excuse “it was T&M.” Trust is lost instantly and rarely returns.

It’s also worth questioning cost-based decision-making where price weighs most. A cheap, poor solution often ends up more expensive. If the budget doesn’t cover a “Ferrari,” a good partner will recommend another sensible option rather than push something unrealistic or unnecessary.

When procuring data or AI projects, what success metrics do you look at and what KPIs would you set?

Success metrics should tie directly to business goals.

Ideally, project success metrics are set at the start. In reality, the more urgent and complex the project, the less time and capacity there is to build a measurement framework. The most effortless situation is when the customer already has set clear business objectives – then project KPIs can be derived directly from them. Common examples include scalability and efficiency KPIs, which can be both qualitative and quantitative. Traditional metrics like schedule and cost come only after that.

One practical KPI is resource requirement: how many internal FTE months the project needs compared to external effort. I’ve seen proposals where internal workload varies widely, and generally, the less internal resourcing needed, the better for the customer.

Since setting metrics isn’t always easy, it’s valuable if the service provider can already propose preliminary metrics, a governance model, a steering group structure, and a cadence for reviewing metrics in the offer. The best players show metrics in early discussions or at least in the proposal phase.

For example, the most mature end-user service providers present clear metrics like customer satisfaction or helpdesk response time and customers rarely need anything above that. The same approach is fully possible in data and AI projects. When a provider can show how they measure their success and link those metrics to the customer’s goals, the collaboration starts on a stronger foundation.

Which AI solutions are mature enough for production, and where should companies still take an experimental approach?

Mature AI solutions are those where the error margin is acceptable and manageable.

At the moment, the best candidates for production use are AI solutions where errors can be tolerated or where an expert reviews the output before it’s used. This way, risk can be controlled and significantly reduced.

Overreliance on AI is also an emerging risk.

One worrying trend is that traditional search engines are used less, and people rely directly on AI generated answers. As AI hallucinations decrease, source citations improve, and text quality rises, these answers are questioned less. This can lead to errors going unnoticed and problems arising that wouldn’t have occurred before – simply because people’s critical thinking fades.

The root cause of errors is often inadequate or outdated data, or misunderstanding what the data should consist of.

Example from everyday life: GenAI might say a restaurant is open when it isn’t, because the source data is inaccurate or outdated. The same risk applies in business – if the source data isn’t current and reliable, the AI will not perform. That’s why data governance and quality assurance are critical before using AI.

In a corporate context, this also means excluding certain data sources entirely. For example, open internet discussion forums are unsuitable as sources because most of the content isn’t fact-based. The quality of GenAI is always directly tied to the quality of the data it’s given.

What challenges are data companies like Recordly particularly well positioned to solve?

Most companies already have long-term tech partners, and there is a constant competition over who gets to be whose strategic partner. Few, however, are true experts in data or generative AI. Customers understand that not every IT provider can be a top expert in every trend – even though in today’s market, everyone wants their own piece of the data and AI cake.

Recordly’s strength is that it can join situations where a partner network already exists and demonstrate expertise without historical dependencies. This enables an objective perspective and focus on how data and AI support the customer’s business.

Specialization is a strategic choice.

In large corporations, group-level management sometimes steer all partners toward the same annual priorities, at times even into unsuitable or conflicting combinations. Recordly doesn’t try to be “everything to everyone”. It specializes in data and AI and can give advice on infrastructure and architecture if needed, but won’t try to compete in hardware installation or basic IT services.

One partner can’t cover everything.

While many organizations want to consolidate all IT services with a single provider, in reality this rarely leads to the best outcome. Recordly is highly specialized, and if the same expertise is requested from a current, larger, and more generalist tech partner, it’s often practically impossible. That’s why Recordly can bring value precisely through its specialization and neutral perspective.

Pete’s favorite quote is by Albert Einstein: “If you can’t explain it simply, you don’t understand it well enough.”