Replacing stock images at Great Learning

The problem with stock

Stock images have become a quiet kind of dishonesty — they’re boring, interchangeable, and the same handful of frames now appear across every ed-tech site on the internet. Since Unsplash made stock photography briefly interesting, the field has cycled right back to images that lack character and any meaningful brand recognition.

When we tried optimising pages for the Indian (domestic) audience, the problem sharpened. Finding Indian faces — let alone Indian contexts — proved frustratingly hard. Ten specific issues kept catching us:

1. Regional and cultural gaps. India has 22 official languages and a thousand realities. Stock libraries don’t.
2. Inconsistent quality. Wildly variable professionalism, resolution, and aesthetic standards.
3. Overused images. We’d spot the same frame on a competitor’s site by Tuesday.
4. Cultural inaccuracy. Misrepresentation of traditions and symbols, sometimes harmless, sometimes insensitive.
5. Limited inclusivity. Minimal representation across gender, profession, and class.
6. Few modern themes. Hardly any images that reflect how Indians actually work in 2023 — Slack, scooters, second coffees.
7. Poor searchability. Weak tagging on stock platforms surfaces the obvious and buries the interesting.
8. Outdated visual styles. Most images fail to align with current design trends.
9. Licensing and cost. Affordable, high-quality Indian-specific libraries are thin and inconsistently licensed.
10. Stereotypes. Over-reliance on saris, samosas, and sandalwood when most of modern Indian life isn’t any of those things.

We flagged this back in 2022. Replacing stock with in-house imagery felt logistically daunting until September that year, when we just decided to do it.

The planning

Before we shot anything, we needed to be specific about what we were capturing — the subjects, the locations, and who each photograph was for. Budgets had to come in close to what we’d otherwise spend on stock libraries, so we kept costs low while pushing hard on quality.

Whom to shoot

Professional models were the obvious answer and the wrong one. They were expensive, the same faces would repeat across scenes, and we cared more about diversity than polish.

We auditioned internal employees instead — real people connected to the brand, photographed in real contexts. We shot across Bangalore and Gurgaon, which gave us a broader pool of comfortable subjects than a single city would have.

For reference, here’s what the old Great Learning homepage looked like before — the stock-image status quo we were trying to replace:

What to shoot

High-quality, relatable scenes — people at desks, on calls, learning, teaching, eating lunch. Things our learners would recognise as their own life. We mapped audiences, wardrobe, and grouping before anyone picked up a camera.

The scenes we wanted to shoot

Every scene was storyboarded before the shoot — references, framing, and intent.

Where to shoot

Paid locations were out. Shooting inside Great Learning’s own offices was also tricky. I called in favours and found an empty office space in Bangalore with the right look. Gurgaon was harder — we ended up renting WeWork rooms and a few alternative office spaces. The locations themselves became part of the storytelling.

Let’s geek out on the camera gear

• Body: Canon 5D Mark IV (full-frame)
• Lenses: 70–200mm f/2.8 (the workhorse), 85mm f/1.4 (portraits, for the bokeh), 135mm f/1.8 (we wished we’d used it more)
• Lights: Elinchrom ELB 400 — a pair — with wireless triggers
• HSS: The flashes weren’t HSS-compatible, but it didn’t matter for indoor work

Art direction

Most people looked uncomfortable at first. By the fourth or fifth shot, they started to relax and act more naturally. The camera is the easy part. Getting someone to forget about it is the whole job.

We were photographers shooting non-actors. The hardest part wasn’t lighting or framing — it was getting subjects to forget the lens was there. Most people stiffened up for the first three or four frames, then settled. I learned a great deal about what makes people anxious in front of a camera, and how much of that is about who’s behind it, not what’s in front of it.

The other thing I’d plan better next time: schedule. We overran on most days, and the cost of Gurgaon location rentals made overruns expensive in a way Bangalore’s borrowed offices weren’t.

Here is how the final images turned out

The frames we kept — real desks, real laptops, real people.

How they got used on Great Learning’s homepage

The final images replaced stock photography across the homepage and key landing pages. The brand stopped looking like every other ed-tech site, and started looking — finally — like a company that knew its audience.