After graduating, I joined a bulge bracket in their European office and have since covered most aspects of TMT but primarily tech.

• Over a decade of experience in IB now: in London, New York and multiple European offices
• Previously led assessment centres and have a very strong understanding of precisely what interviewers seek
• I have conducted 100s of interviews and screened 1000s of CVs

Now MD. Ask me anything---will avoid answering anything doxxing though, sorry. Happy to cover lifestyle, trajectory, networking, hours, conditions, career strategy, ups and downs, etc.

I'll respond via comments and DM.

(also started a consulting prep platform with a friend who's ex-consulting, we'll both be conducting AMAs. We started this subreddit to combine finance and consulting, and will both be answering posts/comments here.)

If your Solve invite says 95 minutes instead of 65 or 85, you're getting the new version of the Sustainable Futures Lab (SFL). I've spent the past few weeks collecting every candidate report I could find on it, and this is where things stand as of late September 2026. If you sat the new build recently, please correct me or add to this.

It looks like the new version is in A/B testing because candidates have said that they got the test, but their friends/cohort didn't. When this is the case, it's tough to know whether the score will be taken into consideration, as they don't typically have enough data yet to calculate your percentile/score in relation to other candidates.

Context:

Solve currently comes in three invite lengths:

• 65 minutes: Redrock (35 min) + Sea Wolf (30 min). No SFL.
• 85 minutes: the above plus the original SFL, the 13-question multiple-choice version, 20 minutes.
• 95 minutes: Redrock + Sea Wolf + the redesigned SFL, roughly 30 minutes.

The 95-minute invite also suggests setting aside an extra 20 minutes for instructions and the technical check, so block out about two hours. Which version you draw looks randomized right now. One September taker polled friends who applied at the same time and was the only one who got the new build. Several candidates also mention a post-assessment survey asking them to review the new format in detail, which reads like a live trial to me. Does the score count? Nobody has said. My advice is to assume it does.

New Version:

The old SFL was a quiz. Thirteen sequential multiple-choice questions in 20 minutes, same linear sequence for every candidate, all wrapped in an environmental-project story. Honestly, it played more like an electronic case questionnaire than a game. The new version is a different genre: a role-play.

You're a project leader with four researchers, each from a different specialty, working out of an island research site with four workstations. The game runs onboarding and then three project days. Every day has the same loop:

• Explore. You review limited information, including which workstations matter that day. It reveals less than you want, on purpose.
• Assign. You place your four researchers. Takers still argue about whether to follow a researcher's stated preference or to cover the day's needs. The ambiguity is the test.
• Support. You check in on your team. A small coin currency buys quick status checks, like a researcher's current workload. My read is that this tests whether you can spend an information budget sensibly under pressure.
• Reflect. You rate how each researcher is doing and how confident you are. Here's the trap. The game never tells you this is coming, so if you didn't track moods from day one, you're guessing by day two.

The two things that blindside people

The diary. Information from a previous day doesn't reliably carry into the next one, and the game never warns you. Takers assumed the diary would persist, skipped note-taking, and got burned. Write everything down the day you learn it. On paper.

The tutorial. It's text-only, and candidates are brutal about it---they think it's rubbish, tough to follow and incoherent. One called it the worst-explained tutorial they'd ever seen, and the less colorful reviews agree that people lose their first 5-10 minutes just decoding the screen. If you walk in already knowing the loop above, that time is yours.

Two smaller warnings. The build is graphically heavy: one candidate's old laptop struggled, and another had to borrow a computer after a graphics check blocked scheduling. And the prompts are wordy. A taker from the earlier version put the pacing rule well: don't reread. You will not get through it if you do. Read once, commit, move on.

What it's testing:

No official word exists, so this is my interpretation. The original SFL scored behavioral traits through multiple choice: prioritization, decision-making under uncertainty, interpreting messy information, balancing trade-offs, and team effectiveness. The redesign keeps the traits and changes the medium. Instead of answering questions about delegation, you delegate, watch workloads, keep people unblocked, and then report on how well you actually read your team. It's a compressed first week of engagement-junior work.

The sentiment backs this up. Takers don't call the content hard. They call it confusing. That distinction matters, because confusion is the cheapest failure mode to fix.

How I'd prepare:

1. Learn the day loop before test day and check for the latest posts RE: the test, as these are coming out daily.
2. Take notes like nothing carries over, because it might not.
3. Track each researcher's state from minute one so the reflect step isn't a guessing game.
4. Read each prompt once, decide, and move. SFL punishes rereading.
5. Redrock and Sea Wolf are unchanged and appear on every invite, so timed practice there still pays off.

Full disclosure: I build prep tools for these assessments. Everything above, plus the underlying reports, is written up properly in our guide to the 2026 Sustainable Futures Lab. Our SFL simulator currently models the original 13-question format, and with the new build this fresh, I haven't seen a faithful simulator of it anywhere yet. If your invite has Redrock and Sea Wolf on it, our McKinsey bundle covers timed practice for both.

Happy to answer questions in the comments.

Hi guys 

I was gonna build a free version of the case partnering thing on PrepLounge. Theirs is nice but it costs $$ after a few cases. 

Is this something people would be interested in/use? Or is it just easier to find a case partner in person and/or making a post somewhere/reddit? 

Curious...