https://ift.tt/ZbWrBUj Ensure your company is working toward Product-Market Fit Photo by Marc-Olivier Jodoin on Unsplash Your Company...
Ensure your company is working toward Product-Market Fit
Your Company is Sinking
As a Data Scientist, you’re mostly trained to focus on data munging and modeling. You’ve got your SQL/Pandas database skills and you can get a nice accuracy score with the right model in scikit, statsmodels, or maybe even EconML. But how does that ladder up to the survival of your company?
A Startup is typically still looking for its Product-Market Fit. Ultimately, that means finding a customer base and creating a product that is so valuable to them that the company’s growth is now unstoppable.
If you’re in the first couple of Data Scientists at a little company of 20–250 people, chances are your company doesn’t have a fantastic Product-Market Fit yet. Once it does, the revenue will skyrocket into $100Ms ARR, your investment rounds will be one right after the other with the valuation tripling each time, and your company will be headed toward 1000 employees within a year.
More likely though, your company is sitting with far less revenue than its burn rate, and the massive MoM exponential growth that you’re hoping for doesn’t quite seem to be coming.
A startup is always sinking. However you look at the reality, that’s a good mentality to have. Until you have indisputable Product-Market Fit, you have an organization that has not yet found the treasure and is running out of time.
What is the Context of your Work?
Your role as the Data Scientist is ultimately to contribute to the company’s Products. That’s true whether you’re directly creating an AI-driven feature, or using data to drive decisions on what should be done with the company’s current and future customers as well as their current and future products.
In order to do the right Data Science AND have your work drive the company in the right direction, you need to know what’s going on:
- How does the company decide on what you should work on?
- How does the company think about the work that you’re doing?
- How does the company decide what to do with your work?
And on and on the cycle goes, as you work with the company to decide what you should build, they look at it, they do something with it (or not), and you figure out what to do next.
If they ask you with the wrong projects, then you’re not fixing the sinking ship. If people don’t understand your work, then, even if you’re able to do something really valuable, it probably won’t end up being correctly understood, prioritized, and utilized.
You Are the Only One that Knows the Data
As the Data Scientist, you’re the expert on Data Science, and probably on math in general. That means that you know better than anyone what can and cannot be done with Data Science at your company.
You know better than anyone what projects are easy or hard, quick or long, cheap or expensive, and hopefully you have a sense of which projects will be fruitful and which ones will not, as well as how your previous projects should be interpreted and utilized.
Is the company fully incorporating all of your knowledge on those questions?
As the person playing with the Data, you have special insight into all the information in it — you’re the one that knows all of the stories that the data tells.
You know what the customers are doing — which kinds of customers are common, and which ones are rare. You know what large swaths of the customers have in common, which use cases are only edge cases, and where there isn’t enough sample size for us to say much at all.
You also know what opportunities exist in the Data. What could be predicted from what? What kinds of suggestions can be made? And again, where is the signal is larger than the noise. If you know what models and solutions are out there, then you know what your company can do with its data — what kinds of products can be built, and what kinds of suggestions can be made — for the company to get the most out of its customers, and for your customers to get the most out of your Products. You know what’s possible in the way of data-driven internal decision making, and in the way of creating data-driven, customer-facing products.
Are you figuring out all of the answers to these questions from your data? And is the company aware of all of the facts and possibilities that you’ve gleaned? You’re the expert. No one knows better than you, and they’ll only know everything if you figure it out and tell them.
Whenever someone does something with Data at your company, the onus is on you to find out, and if it’s not right, to do something about it. Are the analyses as close to causal as they can get? Do the metrics and methods being used actually answer the business question thats being asked? If you don’t make sure, who will? And if you don’t fix it, what will happen?
And Yet, It’s About More Than the Data
All that being said, figuring out what your company should be doing isn’t necessarily about the Data.
For an early stage startup without Product-Market Fit, it isn’t time to focus on minor optimizations.
Until we’ve really built the right product, it isn’t time to look at our churn and make small changes around the customers, processes and features that seem to be related to retention.
Until we have tons of customers feeling like this is the perfect product to them even after they use it for a few months, that it brings insane value worth a massive price, and no one is doubting why they would buy this, it isn’t time to dig into user behaviors to find 10% improvements in the business.
Until you reach that point where your product is a no brainer, it’s still the time to figure out what your company is going to offer. Until then, it’s probably not the right time to work on small improvements to your company’s high cost of acquisition, or its COGS outweighing its revenue, or its high churn. These are simply indicators that your company isn’t offering a good Product. Full stop.
Product? What Product?
So why does this mean you have to be a Product Manager?
The Product organization exists to figure out what Products the company should build, and drive the company toward building, releasing and measuring those products.
What Product is your company building?
It’s the billion dollar question.
It’s the very reason that the business exists: to provide some valuable product. The fact that the business has funding and 100 employees and hired you to start doing some Data Science, does NOT mean that they have figure out what they are actually selling.
Often times, early stage funding is more about the investors’ faith in the team to work towards a future Product-Market Fit. The first little bit of traction, the few paying customers you have, is just a sign that your team was able to find something that a few people would pay for, which is an indicator that your team is capable of finding the right Product in the future — the investors don’t really believe that your company has the golden product now — and your company probably doesn’t have a good enough product to survive.
Until that Product is found, it’s all hands on deck. And if you want to keep your job, then your company has to live. If you want your equity to be worth something, then your company has to thrive. And that means you better do everything in your power to help your company find the right Product.
How to Find the Golden Product
So how do Product Managers find the right product? Well, there are countless things you can do to contribute, so by all means, do some of them:
- Understand your competitors. What are all the companies building something similar to your company: What features do they offer, who is buying them, and why. Do you understand all of these competitors? Are you knowledgable about every other similar and related Product? Is the Product roadmap at your company based on what makes your company special, like what differentiates your company from those competitors as well as what you need to have in common with those competitors (what makes you the right choice for a certain segment of customers)? When you contribute to business decisions, or work on a data-driven feature, does that direction make sense in the context of a market that contains those competitors?
- Understand what your customers want. Your company needs to be doing tons of Customer Research: What problem are you solving? Are you systematically learning structured information about the customers during the Sales and Customer Service processes? Why are your successful clients interested in you, and what went wrong in your failed sales attempts? If you’re a B2C company, then what are the needs, values and feelings that drive customers to your business? If you’re a B2B company, then what are the roles, workflows, processes and strategies of the companies that could be your customers? Do you understand your customer entirely? Are we familiar with every workflow that our customers are experiencing when they make decisions around your Product? What are they thinking when they consider your product versus your competitors? Does our Product fit in the ecosystem with our customer’s other Products (e.g. importing and exporting data between software Products, fitting smoothly into people’s routines/workflows to accomplish their needs, etc.) ? If you know what customers want, then you can push for your company to build the right Products.
- Know how your company works. How is it organized — which team is responsible for what? How does each team decide what it will work on, picking one project idea over another. They may have a prioritization process, or there may be a number of natural, recurring conversations where information and ideas turn into projects. Track it down, figure out how your company chooses what everyone does with their time, and make sure it’s all aligned with finding the right Product. Are they ideating and prioritizing by using everything we know about the market?
- Is your company building things fast enough? Products are like marketing — While we want to be as informed and quantitative as possible, it can still be a bit unpredictable what works. Even though we first must understand our customer’s status quo, beyond that there has to be the innovation of entirely new domains. That happens by being maximally informed and then building a bunch of relevant stuff for customers to try (starting with the most likely winners!) But if your company is only trying a few things per year, then it will probably die before it finds a golden product. It needs to try lots of reasonable new features very quickly. So what is your company spending its time on? Are they fiddling their thumbs debating about small changes rather than trying new things? Are they prematurely optimizing rather than making new products? Are they over-engineering for scale and preventing themselves from rapidly iterating? Are they doing small projects in response to customer requests rather than fully exploring what Products they could possibly offer?
Let’s run through those ideas again:
- Do you understand the other Products in the Market and who they’re for?
- Do you understand your customer’s full experience around your potential Product?
- Do you understand how everyone at company does their jobs?
- Is your company trying tons of great Product ideas?
In sum, is your company doing the right things to acquire this information, and consulting that knowledge while they make every decision?
To me, this all sounds like Entrepreneurship — building a company from scratch. And that’s because you are in an entrepreneurship organization. Don’t be fooled, it’s quite possible that your company does not have a proper product. In reality, it’s a handful of entrepreneurs with nice titles trying to discover a product, with the benefit of a few million dollars and 100 people working on it. So get on their level — be an entrepreneur. Create this business from scratch as if it were your own. Do everything you would do if it were your own company — or at least, make sure that someone is doing it.
Wait, Whose Job Is This?
You might be thinking — hey is this really part of my job? Isn’t this what executives do? Or the Product org? Some research 3rd party? The Sales and Customer Service orgs?
Well, for one, it will always help you make better decisions to know the answers to these questions. If you don’t know the answers, then you’ll propose your feature idea or analysis idea, and either
- Someone will come back to you explaining why that isn’t a good direction at this time, or
- You’ll get the green light to do your project, and then it’s either not used or it’s used and has no real positive impact on the company.
But does the company really depend on me doing this stuff?
Quite possibly.
Sure, at a large company it would be way out of your purview to handle these issues. And at a large company, that’s fine. They only got to that size by buffing the processes enough to consistently capture Product-Market Fit. In that case, these issues are probably handled and you can depend on the chain of command to let you know what to do. Your job there is to specialize.
But if you’re still at a little company that hasn’t found its Product-Market Fit and you want your company to survive, and especially if you want to be valuable enough to move up the ladder one day, then yes, everything is your job. If your company fizzles and you get laid off, feel free to blame whoever you want — but you could’ve done more.
Leadership is Hard
Finally, I have to warn you that your attempts to to what you think is right for the company will not always be well receive. For one, you may of course be wrong. Perhaps you’re missing something. So definitely approach everything with humility. Ask questions, but be humble.
You also have to respect people’s limits. They may not have the bandwidth, in terms of time or emotions, to deal with what you’re bringing. So be strategic — match people’s energy, set expectations, be collaborative.. the usual.
Here are some particular ways that you might try to help your company find the right Product, but you may not find partnerships right away:
- You do what you think is the right Data Science approach to a business question (especially with causal inference) or prototype the right data-driven feature, and other people disagree. They may not understand what you did. They may not see the value because they don’t have the same perspectives on your company’s data, and the market for your products. It may just not be the direction that they were thinking, and the new information takes time to digest. Be sure to present it in the clearest way possible — a document rather than a speech, slides rather than a document, a mockup rather than slides, or a working MVP rather than a mockup.
- Perhaps you think some work needs to be done. Extra information needs to be collected. The system needs to be reworked. The options needs to be explored. It’s probably not the norm for you as the Data Scientist to task others across orgs with new work. Likely, you’re going to have to convince someone that it’s a good idea in order for them to do it — otherwise you’ll have to convince the people around them.
- Finally, if you want to change a process or workflow, you’re really going to need peoples’ buy-in. How they approach their work is personal, changing it is retraining them — it’s a lot more cognitively taxing than doing what they’re used to. Not to mention that it’s an admission that their previous work was less than valuable. And any employee would be unhappy if they disagreed with a process being forced onto them — that’s the heart of much employee disillusionment. Processes can look like natural conversations, how people come up with ideas, how they document what they do, how they make decisions, or really how they do any aspect of their work. If something there needs to change, you’re really going to need their buy-in. Worst case scenario, you’re going to have to convince their boss if the person doesn’t see the value of the change. Perhaps their boss can convince them. The same message coming from a different person can make a big difference.
I think the pattern here is that people resent being asked to do work if they don’t see the value in it. When you bring up issues, you’ve got to come with solutions, because an issue is just more work, but a solution sends us in the right direction, so if the listener agrees, they’ll be happy to walk in a path that will lead to the company’s success. People can tightly close their ears when they’re asked to change or when they disagree, so approach all of it wisely.
But You Have To Do It
In summary, a company only succeeds if its correctly handling most of those crucial considerations about its customers, the market, and what everyone at the company is doing. Make sure that all the intel is being gathered. Create transparency by sharing crucial information, repeatedly with the right people. Help make sure that all this information is properly used at each decision, from the top level company strategy, to all of the company’s processes, to every project that everyone is tasked with.
Because your company needs to figure out what it should be making.
Or else, it dies.
At a Startup, a Data Scientist must also be a Product Manager, and more. was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.
from Towards Data Science - Medium https://ift.tt/QrY49O2
via RiYo Analytics
ليست هناك تعليقات