The research industry has, understandably, never been more excited about innovation.
AI, automation and new technologies are changing how we work and solving genuine
challenges. However, for all the excitement around what’s new, projects still succeed or fail for surprisingly old-fashioned reasons.

What do I mean by this? Well, it comes back to the basics – ownership, communication, accountability, realistic timelines and quality assurance.

This is not a criticism of technology (far from it). Some of the best developments in our
industry – every industry, actually – have come from innovation. Fraud prevention tools
are helping us tackle increasingly sophisticated respondent fraud. Automation is
removing repetitive tasks. AI tools are improving efficiency workflows.

As I see it, the problem is when innovation becomes the objective rather than the
solution. When it is trying to make up for operational lacking elsewhere. When we look
to innovate operations (ops), the question should be “how does this improve delivery?”
– because technology can enhance great operations, but it cannot replace the
fundamentals of great delivery.

Great operations often go unnoticed

One of the challenges with operations is that nobody really notices it when it’s working
well. Clients rarely see the risk logs, quality checks, escalation processes or project
reviews taking place behind the scenes. They don’t see the conversations happening
internally to identify risks before they become issues. Instead, they simply experience a project that feels under control.

They know what’s happening. Communication is clear. Risks are raised early. Issues are
managed before they become crises. The project progresses as expected. In many
ways, great ops is often invisible when it works. When it doesn’t, everyone notices.

Where projects really succeed or fail

Technology can support delivery, but ultimately people still need to own outcomes.
Someone needs to make decisions. Someone needs to escalate concerns. Someone
needs to take responsibility when challenges arise.

In my experience, good ops does not mean that nothing ever goes wrong; research
projects have too many moving parts for that to be realistic. The difference is how early
you spot a problem, who takes ownership, how you communicate it and whether you
stop it becoming something bigger.

We often use a traffic-light system to track project health, and I think there’s an
important difference between a red project and a silent red. A red project has a known
issue that has been identified, communicated and is being actively managed. A silent
red is far more dangerous. The issue is there, but it hasn’t been raised. By the time it
surfaces, the options available to solve it are often much more limited.

Operational excellence is about recognising risks early and responding appropriately.
You could certainly say that proactive risk management is a key pillar of good ops.

Data quality should be built in from day zero

Another misconception is that quality happens at the end of a project, but that’s not the
case at all. Quality is never – or at least, should never be – created during a final review
or a last-minute check before delivery. It should be built in throughout the entire project
lifecycle.

You cannot QA quality into a project at the end. If feasibility was unrealistic, the
questionnaire introduced bias, routing wasn’t properly tested, poor sample entered
field or warning signs were ignored during fieldwork, a final data check cannot undo all
of that. Every decision contributes to the quality of that ultimate result.

That’s why data quality should never be viewed as a single checkpoint. It is the
cumulative result of myriad decisions, checks and processes all working together.

The commercial impact

Poor ops affects the commercial performance of a business just as much as it can
affect delivery. Rework costs money. Delays cost money. Unmanaged scope costs
money. Poor supplier decisions cost money, and fixing something late almost always
costs more than dealing with it early. Beyond the immediate financial impact, there
is something potentially more damaging: loss of trust.

Ultimately, clients are looking for reliability, and reliability is impossible without strong
operational foundations. Poor communication can quickly erode client confidence.

Once that confidence disappears, so can the likelihood of repeat business. Operational
excellence, governance and rigour aren’t administrative exercises. Done well, they
create commercial advantage. Done poorly, they become a liability. Good operations
does more than simply protect delivery, because it also protects margin, client
confidence and the next piece of work.

What happens when things go wrong?

Operational excellence is sometimes confused with perfection. In reality, problems are
inevitable. What matters is what happens next. How quickly does the team act? Who
takes ownership? Is the client kept informed? Do we fix the immediate problem, or
actually understand why it happened?

In operations, ownership counts for a lot. Sometimes a well-managed recovery can
actually strengthen a client relationship, not because the problem was acceptable, but
because the response demonstrated honesty, control and accountability.

Taking it back to basics

Innovation will continue to change research, and that’s exciting. We should embrace it.
But we shouldn’t confuse having the newest technology with having the strongest
operations. Clients will remember whether we delivered. Whether they trusted the
data, whether we communicated when things got difficult and whether we did what we said we would do.

Technology will keep changing. Those fundamentals won’t. Perhaps that’s the
irony. In an industry understandably obsessed with what’s next, some of the things that
matter most are still remarkably old-fashioned.

Payal Jani is operations director at Research Clever