top of page

We built and discovered the problem of talent ‘freshness’

Talent freshness is the issue of candidate relevance to the job market — i.e. not just that a candidate had been looking for a role, but how ‘fresh’ is their desire to find a new role.

This is about our experiences building the talent database platform, and what we learnt about talent-freshness in the HR/recruitment process.

A side project,, started during COVID in April 2020, is a talent database that laid-off talent could use to list themselves and their backgrounds, with an option for companies to also post jobs. We started this initiative to help friends of ours who were looking for new jobs in the middle of COVID.

Started by Harry Schiff, Liu Simin, Rachael De Foe, and myself over a weekend, we managed to get 27 VC funds on board supporting the initiative, list 1,500 individual awesome people, list 500 jobs from startups still hiring, and although we couldn’t track who ultimately got a job through the database, received great feedback from some of those who did — with some receiving interviews in 24 hours.

We were featured by various regional and international media — somehow also making it all the way to Japanese media.

Our origin story:

The story of how we built SEAcosystem came from the fact that many of our friends were looking for roles during COVID, and we figured that there should be a better way to brainstorming and making referrals.

Over a weekend, we quickly hacked together a Google Sheet and Talent Submission process, inspired by a similar concept started by Floodgate, Unshackled VC and Awesome People in the US.

We envisioned this to be a short-term project that might be helpful to a few people, and that would probably fizzle out in a few weeks. 6 months later, it is still an actively used sheet that has been used by thousands.

Our story and experiences are mostly focused around relatively high-skilled white-collar workers working in the tech/startup space, but I believe these lessons can be generalizable to other sectors of talent.


What we learnt:

In a traditional job market, the majority of top talent is headhunted (inbound search) rather than outbound.

During COVID, this dynamic was turned on its head, with more talent being laid off and entering the job market then was traditionally expected.

This flipped the traditional dynamics — with candidates feeling more pressure to publicly signal to the market that they are open to hire (i.e. outbound from the candidate).

What made SEAcosystem unique was its ability to allow genuine candidates to signal that they were looking for a job since there was friction in the application [candidates had to fill out a list of information], the candidate was genuinely searching for a job at that moment.

When it was launched, it was the only platform for full-time tech/startup talent where recruiters could confidently say that all leads were available for immediate start.


Many platforms allow for discreet signalling that one is looking for a job. LinkedIn allows you to signal to recruiters that you are discreetly open to work. The issue with this is ‘freshness’ — candidates may leave their profile settings as ‘Open to recruitment’ even after finding a job due to oversight, lack of need, or simply as a means of increasing optionality without a genuine intention to change their job. There is no impetus or urgency to ‘close’ your open status.

Eventually, one of the product features that was released by LinkedIn included the tag ‘Open to work’ within your profile picture. This is likely to create some urgency to ‘close’ one’s open status, as when one does find a job, they are unlikely to keep a profile picture with the ‘Open to work’ tag.

An example of the ‘Open to work’ tag

It does not appear to be a feature yet, but it would be ideal to be able to search for profiles based on the tag that appeared within the profile picture.


Questions to take from this experience:

There appeared two key opportunities to solve:

Candidates entering the market: How do you create a shortlist of candidates that are looking for jobs on a just-in-time basis?

We already know the tech exists to predict employee retention — can it be used for proactively poaching too?

Furthermore, many talent lead generation platforms/agencies already aim to increase the LTV of their candidates (i.e. keeping track of successfully placed candidates in view for their subsequent positions thereafter). This is particularly true for high-demand, high churn employees (Software engineering for e.g).

It would be interesting to see if there is an opportunity for third-party talent prediction tools — the low hanging fruit seems to be that rule-based search capabilities (rather than jumping straight into “machine learning”) may be powerful enough to proactively track relevant candidates.

Think MySQL for candidates with an RSS trigger

E.g. when I’m searching for talent for a junior-mid stage strategy role:

  • Notify me when MBB consultants are 1.5 years into their MBB job

  • Notify me when any candidates with the term {‘strategy’} in their {Job Title}, {2 to 4 years} out of university in {X Region} post content with the following triggers {‘#opentowork’}

It would be interesting to see both real-time trends of the types of profiles that are taking on certain job scopes, as well as data-driven insight into and tracking of the type of profiles that should be targeted for specific job descriptions

Platforms like Hiretual have recently emerged, with an emphasis on multi-platform data, hinting at the difficulty third-party tools had with integrating with LinkedIn.

Candidates exiting the market: How does a platform solve for when a candidate is no longer in the market?

To solve for candidate relevancy, closed-loop ecosystems are probably necessary to efficiently get the data to close off ‘dead leads’, allowing companies and recruiters to more confidently focus on targeting leads.

This emerged as a problem for over time. Although we removed hundreds of profiles that eventually found a job, believing that individuals would not want their details in a public job search portal, and encouraging candidates to write in once they had found a job, we still received feedback that listed candidates were not in fact active in the job market and had already found new jobs.

This seems to be a problem that can only be solved with the same platform controlling both the distribution of available candidates and the closure of the job itself (so they have visibility on which candidate received the job and is therefore unavailable).

It would be interesting to see if there is an opportunity (and if it even makes business sense) for a third-party solution that helped talent-first platforms (as opposed to job-first platforms) remove ‘dead leads’. My suspicion is that closing the loop, although a pain point, is not a significant enough problem that can be solved by third-party solutions.

Nevertheless, it is clear there is a massive scope for automation in an industry with relatively low barriers to entry. The recruitment agency is highly fragmented, and the market size is huge. It is almost inevitable that the verticalization of the talent market would result in LinkedIn’s disruption and unbundling, creating separate talent ecosystems across different industries. As that occurs, third party solutions, especially around search and tracking will likely be more and more useful and necessary.


Thanks to Daniel Callaghan [Veremark], Prerna Sharma [Antler] for their thoughts/comments on the article.


Couldn’t Load Comments
It looks like there was a technical problem. Try reconnecting or refreshing the page.
bottom of page