Why DeepSeek’s Talent Strategy is Core to Its Disruption

刻舟求剑: Why DeepSeek’s Talent Strategy is Core to Its Disruption

Yes, this is my first time using Chinese in one of my article headers. Working in both languages, I have always said that English was a bit lackluster in describing an illustrative idea using few words. The title of today’s article literally translates into “marking the boat to find the sword.” If you’re curious on the story and meaning behind it, check it out here.

Okay, back to regular programming: Ever since DeepSeeks released R1 a few weeks ago and caused some waves in the industry, I have been really curious. Namely, how can a company become so disruptive so quickly in the geopolitical and economic environment it was created in, with all of the limitations and restrictions on technology access?

Needless to say, I delve down another rabbit hole. This rabbit hole required more digging than usual because the founder of DeepSeek, Liang Wenfeng, hasn’t done an in-depth interview with any Western media. So, of course, I had to dig through Chinese sources for materials. While tedious, this little research endeavor yielded more than I had hoped for.

In reading all of the interviews with Liang on the Chinese version of 36Kr, I realized that DeepSeek’s unique approach to talent and culture could well explain why they have become the disruptor in the market. Now, for those of us who have worked in HR for a while, nothing DeepSeek did was new. They just had the persistence to keep at it for a few years until the results spoke for themselves.

Before I get into this, for context: Liang graduated from a domestic Chinese university. When most of his classmates went to work for big names like Tencent, Bytedance, and Alibaba, he decided to pursue AI. This was back in the late 2000s, when AI wasn’t that big of a thing outside of research circles. Before DeepSeek, Liang ran a hedge fund company called High-Flyer Quantitative Investments and turned down a partnership role at DJI (yes, that one).

I share all this because it sets a really important context for the type of business leadership and vision needed to develop a workforce that can disrupt a whole industry.

So, what did DeepSeek do exactly? Here are the three key highlights:

1. Ignored Experience

In one of his interviews from 2023, Liang shared that instead of hiring for experience, DeepSeek looks for someone who has the foundational skills, is a creative thinker, is deeply curious, and more than anything else, has the passion for DeepSeek’s mission. Paraphrasing a few quotes from that interview, we can start to understand DeepSeek’s different approach to talent. According to Liang:

• If you are after a short-term goal, then hiring for experience makes sense. But if you are working on something that requires a longer time horizon, then experience isn’t really necessary

• You don't have to have done something previously to be able to do something in the future

• Someone with experience will tell you exactly how to do something. Someone without experience will repeatedly experiment, think about the problem thoroughly, and come up with a solution more fitting to your current circumstances

Putting his money where his mouth is, most of DeepSeek’s founding team were new grads, those who graduated 1-2 years ago, and sometimes even PhD interns. DeepSeek also skipped something in its hiring approach that most Chinese tech companies go after: they did not hire any “Sea Turtles 海归” (it’s a term used for returning Chinese expats who studied overseas and are seen as having more exposure to best practices, etc. from their experiences abroad).

Now, not hiring for experience sounds easy, right? After all, we’ve been discussing not putting degrees as requirements in job descriptions for a while now. So, if that’s the key differentiator about DeepSeek, why are we not seeing more disruptors across industries?

2. Cultivated Culture

Talent development doesn’t happen in an instant. When asked about whether he was worried about his competitors replicating his hiring strategy or poaching his team, Liang said that his company’s talent pool isn’t the only element here, as strong talent requires a unique company culture and management philosophy to go with it.

For example, two key sales roles at DeepSeek are occupied by someone with a background in importing/exporting German-made machinery and someone who used to be a backend developer. To support them in their sales roles, the DeepSeek management team had to look at performance KPIs differently. In fact, neither of the salespeople did any deals in their first year (unheard of by current account management standards), and it was only in the second year that they started to see movement and revenue from the market.

According to Liang, DeepSeek’s approach to talent development is to hire smart people, give them tasks of value, not tell them what to do, and let them come up with their own solutions. Creativity requires minimal management and intervention; it has to be organic and not forced.

Instead of documenting culture in a playbook, DeepSeek opted more for the management-led approach, whereby the management team's actions set standards and norms for the organization, and they lead by doing.

3. Threw Out the Rulebook

One thing that came out loud and clear to me after reading through the interviews is that DeepSeek doesn’t follow a formula or rulebook. Early in one of his interviews, when asked how he feels about competition from larger Chinese companies, Liang said that while larger companies have a size advantage, they are also slow to take something from theory to application. Due to their need to see results fast, they also might not have the persistence needed to see an innovation through to fruition.

To me, part of DeepSeek’s success can be attributed to the fact that they understood the rapidly shifting market conditions basically rendered all previous experiences and practices useless. According to Liang:

• The market is changing, and success isn’t based on rules and processes. It is more about adaptability and flexibility

• Most companies are too easily restricted by their own experience and habit developed overtime

• The difference between DeepSeek and its competitors is not a time difference of 1-2 years. It is the difference between originality and replication

The idea that inspired me to write this whole article is from an interview back in 2024 where Liang said: You cannot use the business sense from the internet era when you are thinking about how to monetize in the AI era.

Just like how you should mark a boat to find a lost item in the river, sticking to the same things we have always done may be why HR practitioners are experiencing budget cuts and feeling less relevant in their organizations.

If you just scrolled through the bulk of the article, take away this one thing: use DeepSeek as an example the next time you are being questioned if a new talent management plan or culture transformation will work. DeepSeek is proof that it will work as long as you are willing to put the work behind it, consistently, and across the organization.

As I mentioned at the start of the article, none of these three things are new or groundbreaking in any way. As we stand at the start of the AI era, there are two things we should all think about: 1) This is different, so what will you do differently? And 2) seeing IRL that the talent practices we’ve been talking about for ages actually work in practice, how will you find a leadership team (or convince your leadership team) that it’s time to change how we’ve always done things in talent management?