A Hitchhikers Guide to Tech Business Interviews

This guide is designed to help candidates seeking business roles (Strategy & Operations, BizOps, Finance, Corporate Strategy, Growth, Marketing, Operations, Analytics, etc., excluding PM — outside my area of expertise) in the technology industry. It can also be generalized to some degree to other industries.

This guide is based on my personal experience with companies ranging from Series A/B startups to FAANG and everything in between, and I wanted to share my learnings with anyone who may find them helpful in their career search.

Interview Process Outline

The “standard” format for a recruiting process looks something like this:

1. Recruiter Screen — straightforward, usually just a walkthrough of your experience and interest in the company + logistics (e.g., salary requirements, location, work authorization). Some will have light behavioral STAR (situation, task, action, result) type questions, but nothing too in-depth.
2. Hiring Manager Screen — A bit more involved, typically involves telling your story/background, some STAR questions, and potentially a short case.
3. Skills Assessment (usually take-home case but can also be live) — very involved, will discuss more in-depth in the following section.
4. “Onsite” (but 90% are virtual )— usually presentation of case with additional mini-cases, will discuss a bit more in-depth in the following section.
5. Executive Interview — sometimes encountered, typically an interview with the “skip” who is either a VP or C-level to gauge higher-level motivations and thinking. Can also be more of a “sell” call at this point too where they try to convince you to take the job.

“Standard” is probably a bit too generous for describing the format, companies can do this in any mixed order or, in some instances, have multiple cases or other assessments in the process (e.g., an IQ-type test, live modeling, SQL coding, etc.). So be ready for anything and everything.

Skills Assessments

Assessments can be broadly placed into six buckets:

1. Data Analysis — Get a raw data set (usually 1k to 50k rows) in CSV/Excel related to their business in which you are asked to find insights and make recommendations in either a slide deck (consulting-style) or written document (Amazon style). Excel/Sheets are usually the right analytical tool for this with Pivot Tables being the most important function/skill to apply to aggregate data and get insights. I prefer Excel since the pivot table/charts functionality in Sheets just doesn’t cut it for me. I’ve also used Tableau (better visualization and ability to do things like median) and Python on occasion, but it’s usually super overkill for these things and should only be used if you’re 1) comfortable with it and 2) the position/hiring team has given indications that they use it. Not a good use of time to learn these things on the fly but if you know them and you know the team uses them, can be bonus points to help you stand out (and on the flip side, if you know the hiring manager doesn’t code…using Python may not be ideal). Will go into more detail on this after the break.
2. Modeling — Get some one of or some combination of : 1) baseline assumptions in written format (e.g., churn rate), 2) a historical P&L with broad categories (e.g., revenue, COGS, R&D,), 3) raw data similar to the data analysis exercise. The goal is usually to build some sort of forecast (e.g., forecasted P&L for finance, forecasted sales capacity planning for sales strategy). This tests for a different muscle than data analysis, you have to be able to make assumptions based on limited data, defend them if asked in a presentation, and then use the assumptions to predict the future. Excel/Sheets + PowerPoint/Slides (sometimes) are the only tools you’ll ever need for this.
3. Open Ended — Get an open ended prompt that is typically around building a business plan from scratch. Most commonly found in consumer marketplace businesses for GM/market launcher roles for earlier stage companies (more mature companies do #1) — an example fictional prompt would be “Design a plan for launching houseplant sharing in the Minneapolis market”. Keys to success are creativeness in demand generational/marketing and thoroughness in operational implementation. Can sometimes choose between slide deck or written document — know your audience, if it’s ex-Amazon you’ll be presenting to, lean doc, if it’s ex-consulting then lean slides. I really enjoy doing these.
4. Onboarding Plan — Asked to map out what you would do in your first X days at the company in the role. Slide deck or doc deliverable. An uncommon exercise that’s usually associated with early-stage companies, larger companies already have an onboarding plan for you so there’s no need for this. Tip — early-stage companies move super fast so you need to map out concrete deliverables that you’re going to bring to the table within a short-time frame, you cannot just put training/learning type activities that would be common for the first few months at larger companies.
5. Excel Competency — Get a multiple choice-type timed test based on an CSV/Excel data dump that has actual right/wrong answers. This is a straightforward exercise that tests for basic Excel knowledge such as pivot tables and VLOOKUP. An example question for an e-commerce store is “how many blue pans were sold in Texas for the month of August”. I don’t come across these that often — IMO they are pretty pointless and as a hiring manager I would never do them. I get the idea — it’s a low-effort weed-out mechanism to filter people who can’t do Excel, but you learn nothing about a candidates critical thinking skills and ability to deal with ambiguity.
6. SQL Coding — SQL test in either an online tool such as Hackerrank OR 1:1 problem solving/coding session with an interviewer. This is fairly uncommon for most business orientated roles in my experience — even if they technically require or strongly prefer SQL knowledge. For analytics/data science roles, 100% you’re going to get one. If a company asks for one in a business role and you don’t know SQL or don’t know it very well, the test is likely easy enough where one week of studying should get you up to speed (Mode’s SQL Tutorial is fantastic). Think Leetcode easy-type questions that test for basic select, join, and aggregation functions — you may also get asked to explain some of these. concepts. For analytics/data science roles, usually it’s quite a bit more involved and moves on to Leetcode medium/hard and will test for things like subqueries, window functions, null handling, etc.. in additional to more advanced conceptual things such as indexing and partitioning. If you’re interviewing for this and aren’t very familiar with SQL — IMO it’s probably the wrong role and it’s very tough to get to this level of competency in a week.

While one singular assessment is most typical, some companies may ask for more than one type of assessment in different stages or in one combined stage. Should you reach this stage, there is typically an opportunity to present to a panel, but a few companies may reject you without the opportunity to present if they don’t like what they get (has happened to me once).

Assessments are typically take-home “homework” with a reasonable lead-time (enough to give you a weekend to work on it), but some can be time-boxed (e.g., 1 hour window to do and send back) while others are “live” and require you to complete and present in a 2–3 hour block. I prefer the latter “timed” versions so the work is not weighing on your mind for days and you can knock it out quickly, but they are intense pressure tests, and you can easily mess something up with limited time to recover or just have a bad day.

For take-home assessments, companies will often tell you that you should spend no more than X to Y hours completing the assignment. That’s usually never enough time, IMO — if you want to succeed, plan to budget at least one working day (8–12 hours) of “heads down” time to do it.

Data Analysis Deep Dive

Since #1 data analysis is the most commonly used form of assessment (~50% of interviews in my experience), I’ll do a further exploration of “what good looks like” here.

Raw datasets are typically just one tab/file but sometimes they can be separated into multiple sheets on an Excel. If it’s multiple tabs, usually one tab contains “fact” data (e.g., e-commerce transactions) and the other contains dimensions (e.g., e-commerce customer characteristics)—all you need to do here is join the sheets into one dataset using VLOOKUP or similar so you get everything in one sheet.

Once your data is in order, look through the columns to make sure you understand them, and then get to work on exploratory analysis. This is my personal process:

1. Identify key KPIs/metrics (counts, ratios, etc.)
2. Slice and dice the data by dimensions (e.g., characteristics such as time, geography, category) and find interesting trends (pivot tables are great for this) for said KPIs/metrics. Keep track of them — once you’ve found something interesting, lock it down by doing further analysis on a new tab you’ve copied.
3. Always be generating hypothesis in your analysis!!! There’s two approaches to this — you can either start with a hypothesis (e.g., maybe sales are higher on weekends due to people being at home) and then see if the data validates the hypothesis OR slice the data, see an interesting insight, and come up with a hypothesis of why X is happening (e.g., noticing an interesting trend then thinking why that may be). Doing both approaches simultaneously is best IMO to generate the most comprehensive set of insights.

Now that you have insights, it’s time to put them into a presentation. I’m going to use a slide deck example here, but this could easily be turned into doc format as well.

Here is an example of a slide for a hypothetical food delivery company.

Fictional food delivery slide — entirely fake data I made up

I use this two-column slide template for every one of my take-home cases, and it’s been successful. Having a standardized slide template is extremely helpful because it eliminates having to think about slide design. Some of my successful presentations are literally this slide repeated 8x with different content.

I would advise not wasting any time on fancy slides with icons and cool visuals that you might use in consulting or even aligning boxes. Basic is fine; what is important are the insights and recommendations. The only exception here is if the hiring manager is/was a career consultant, in which case it might be worth it to throw a little bit of extra jazz into the slides.

The framework I use here is VIHR (Visualization, Insight, Hypothesis, Recommendation)—don't try to Google this; I just made it up.

1. Visualization — Usually a bar chart showing the interesting trend that is happening. You can also paste in pivot tables here as well, but I find graphs are often better to get the point across.
2. Insight — Explanation of the insight that you see on the chart. This should go beyond just describing the trend, it should also include the potential implications of the trend. Don’t tell me the sky is blue, tell me why that matters!
3. Hypothesis — A theory on why the thing you are seeing is happening. Sometimes you can get this from the data, sometimes not, in which case you should still list out potential reasons. This is important for driving the recommendation.
4. Recommendation — These are ideas to improve the business based on the insight + hypothesis above. The more ideas the better. At this stage, it doesn’t have to incorporate a more detailed implementation plan. If X is happening, I suggest we do Y.

When writing slides and presenting, it is important to be confident in your analysis and recommendations. Treat your audience as your client. Do not add question marks (?) in the slides or add implicit ones in the oral presentation (upspeak, making a statement into a question).

To wrap it up, it’s also helpful to have an executive summary as the first slide summarizing all your findings and recommendations.

Onsite

This involves a series of 1:1 interviews that take 2–5 hours. If you completed an assessment, you will likely have the opportunity to present in front of a panel and field questions.

Companies typically take two different directions in the onsite when it comes to who you will meet with:

1. Team-Only — you only meet with members of the immediate team that you will be working with (e.g., Bizops). Much more common if it’s a generalist Bizops role where you may be assigned to different projects that are not yet known.
2. Cross-Functional (XFN)— you meet with people on your team and cross-functional partners that you will work with. I much prefer this format as you get a better sense of who you are working with, but it can be harder for companies to calibrate and schedule.

I won’t go too much into the behavioral/STAR type questions here, you can find standard advice on this elsewhere — just prepare 8–10 stories that you can modify to fit most situations.

Instead, I’ll go over the types of potential cases that you may encounter during this stage. The majority of companies will do cases, however a decent minority may choose to have these be entirely behavioral/culture fit. Either way, you should be prepared.

1. Company Case — a business case on a situation that the company or interviewer has encountered before for the company. Usually it’s very similar to a consulting case, but without the calculation part, less structured, and always being about the company you’re interviewing for (not something random). You will have to come up with a framework for approaching the problem and brainstorm ideas. It may sometimes include a brief market-sizing exercise. This is the most common format.
2. Product Light — a product case, typically about metrics. You can google PM interviews to get a sense of what this looks like. It usually isn’t as extensive as a full PM loop (e.g., no product design), but it’s similar. If you’re interviewing with a PM this is the likely format.
3. Presentation — presenting your take-home assignment to a panel of interviewers who will then drill you with questions both during and after the end. Be confident in your presentation but also keep an open mind for other ideas/challenges you may not have thought of.
4. Consulting Case — a business case pulled from a consulting case book that is not related to the company. This is rare, but happens. I find this practice rather confusing and unhelpful for either the candidate or the company.
5. Personal Case — some case that the hiring manager makes up from their experience (or completely fictional).
6. Brain Teasers — logic and probability/combinatorics questions (e.g., dice rolls) that are completely unrelated to the company. Most companies have gotten away from this type of thing, but it still happens.

To prepare for these interviews, I typically do three things:

1. Online research — read as much as you can about the company. This can include TechCrunch, Reddit, WSJ, 10Ks (for public companies) and more. Google is your best friend. Since these are business roles, you should always thinking about how the company makes money, for some companies it isn’t super straightforward. One of the things I like to do is jot down ideas on how to improve the company/product, this will sometimes come up.
2. Experience research — if you can, you should sign up to be a user of a product and try it out. For marketplaces, it can also be helpful to try and understand both sides of the market. Everyone has ridden an Uber but few candidates have actually driven for them. This is a bit harder for SaaS (especially enterprise vs. PLG-type products), but if the opportunity is there sign up for the 14-day free trial.
3. Networking — if you know someone or have a 2nd degree connection that works/worked at the company, it can be helpful to reach out for a quick chat. You can learn more about the business from a more casual perspective — especially helpful if the business is not a mainstream/household name (e.g., Doordash). I don’t do this too often since it does take some time to setup, but the few times I’ve done it, it has been helpful.

Other things that some people may do to prepare include practicing random consulting case interviews, using Rocketblocks, or practicing interviewing with a friend. The first two can be helpful if you’re interviewing for a consulting-type role often full of ex-consultants and you’ve never been a consultant/been through the consulting interview process. If you have, you can typically get by on muscle memory alone.

Bonus Section — Generative AI

I’ve never used these things for any of my interviews but now that Chat GPT is in the wild I was curious to understand how I could potentially leverage it (and other tools like it) to help with case studies. So I tried this with the example I had above on food delivery.

The result is pretty stellar for this particular example. It isn’t perfect (#6 makes no sense as these services don’t have “fulfillment centers” and #5 is strangely non-specific) and the responses are a bit generic, but it does throw good ideas in volume so you can pick out the best ones and refine them (and add your own). This is just the FIRST 7, you can ask it to continue and it will give you more ideas! You can also review the answers and tell it that it is wrong. It is able to be much more “collectively exhaustive” than almost all humans. I haven’t tried more esoteric examples on it — but you can try.

While Chat GPT can’t do the analysis for you yet to find the trend and put it into context (Excel can kinda do the former but not the context part, it will happen quickly though!) — if you ask it for ideas from a good prompt then you can get decent results.

Some people may consider using Chat GPT for assignments “cheating”, but when I thought about it for a bit, I would disagree. The point of these interviews is to understand how you would work →the future of our work will include tools such as Chat GPT and it will be just as ubiquitous as Google. So if you can use it to make your work better, that’s good for the company as it will lead to more profits. The ethical lines are a bit more blurred for something like academics, but for this? Go to town.

Conclusion

I hope you found all the information above helpful. The tech downturn/layoffs right now is brutal and is impacting a lot of people’s lives in ways I can’t even imagine, if you’re in a tough spot and looking for a new role, if there’s any way I can be of help please don’t hesitate to reach out.