Reviewer’s Note: This feedback is coming from the experience as an hiring employer as well as job-seeking candidate on AngelList – but not one of an investor nor a fundraiser. AngelList was pivotal in introducing the concept of crowd-based funding rounds, and now investor syndication, areas that I don’t delve into much (yet). Plenty has been written on both this concept as well as Naval’s efforts to change laws around crowdfunding. Initially labeled ‘Angel List Talent’, the now incorporated jobs / recruitment functionality of AngelList has been a happy side effect to the greater forces at play on this platform.
The fervor around hiring Engineers and Developers around San Francisco has grown to a pitch that hasn’t been seen in my lifetime – not even pre-crash of 2001. Seemingly everyone has both concepts and capital, but no ability to execute – forcing even tech giants to collude (illegally) to prevent poaching from one another. This demand has spawned schools like Hack Reactor and General Assembly to rapidly churn out supply – graduating ‘fresh to SF’ classes of extremely specialized front-end coders, trained in 12-week intensives on whatever the most relevant elements of a changing technology stack might happen to be. The placement rate of these newly crowned, with their 18k in newly incurred debt, 3-months vocational training but often minimal professional experience consistently sits above 95%, with the average salary of 100k for a junior position. Referral bonuses of $30k are now being seen, offered to employees who can find and place engineering (and design!) talent successfully internally. Come on in, the water is fine.
So, if you have even a glimmer of a professional network, your evenings are best spent finding, understanding and reaching out to the ever-so-rare, Northern California neck-bearded polyglot, the full-stack engineer of yore. They’re out there – it is just hard to find them. Yet, the primary search and recruiting tool in used in our (and any technology) industry is a trainwreck of an experience. LinkedIn, only my second foray into social networks (I suspect for millions of others as well), registering a mere six months after initially signing up for The Facebook back in 2007. Hampered by it’s own phenomenal growth, LinkedIn has now become bloatware to the point of being unusable and impairing the very connections between professionals that it was created to foster. The content of this article goes deeper into some of the differentiating factors between the user’s experience of using LinkedIn versus AngelList as a networking and a recruiting tool. I go on to reason why a private company of less than 50 employees will eventually overtake a site that has 300,000,000 members with a valuation of $24B.
Missed Opportunities & Lost Platforms
The initial experience of LinkedIn offered great promise. Creating one’s professional network was a natural extension on how were already learning to craft our virtual personas – still a new experience for the early adopters. With a virtual tear streaming down my cheek, I still remember my first LinkedIn connection. Now, as I’ve accepted my something like a 943rd connection, the most utility I find through a LinkedIn experience is a place that where my long-since retired father and I have can publicize snarky Dilbert cartoons. It is always a good indicator on the desaturation of relevancy for any given a social network when the preceding generation has begun to use it as a medium of communication. Facebook past that point long ago.
Daddy issues aside, my activity on LinkedIn mirrored its extensibility. Initially, LinkedIn encouraged relationships with their content partners and third-party developers, offering more holistic and eclectic view member profiles. What initially started as a wonderful platform of extensibility has since been removed. No longer will your reading list and book reviews from Amazon port over (to show that you are actively reading industry), nor will any shared TripIt actions (to show that you are attending conferences and furthering domain knowledge) appear as components of your profile. Several years back, LinkedIn made a bold mode and killed any integration with external applications, sealing off the inbound gates and deciding that a closed loop system was how they wanted to control the experience. I can only guess that this assumption around their product strategy was based on the notion that no one should need nor want to consult anything but LinkedIn for any information on their fellow professionals. Control.
Need a PDF of a resume? Nope (nor is there a media-type query for when I try to print a profile page). Want to find a historical record of an connection’s status updates? Nope. Want to showcase a GitHub repo / StackOverflow profile? Nope. And what the fuck is WeChat? The hidden placement of these other crucial, but now obscured parts of our professional personas are hidden behind a collapsed component – for fear that we start considering other options besides LinkedIn for reference, or even the convenience of exporting data. But, there is no shortage of functionality for LinkedIn – they’re a powerhouse for feature releases, seemingly all the wrong ones in nothing is being removed nor rethought.
Corruption of a Community
Any social network is full of scud accounts. It is estimated that Twitter, where anyone can follow anyone else, has an estimated 1-in-10 fake followers. Facebook certainly has it’s share of fakes, although with only connections created through user-approved actions, this lowers the rate somewhat. LinkedIn — although they instruct their members to only connect with people they know — there’s no longer that barrier in place if one pays for a Premium account to remove that barrier of access.
The business professional on LinkedIn wants share knowledge and achievements, create meaningful connections and maintain them. The Recruiter wants to achieve, by volume, as many eyeballs as possible for their open positions. External recruitment without attachment to a company is a false community, in the fact that they’re witnessing information exchange and contributing very little but the generation of spam. Just ask Rupert Murdog.
The recruiter does not have to abide by any rule that could possibly cause detriment to the company they represent (i.e, message spam) they’re just playing the numbers. Following the same Salesperson / Business model they position their user accounts around, LinkedIn has enabled Recruiters to do take these actions, by granting them access to purchase InMail credits and superseding the same prior knowledge restrictions they caution the user againstl. As Facebook has profited from businesses posting advertisements within your FriendFeed, LinkedIn has continues to push the monetization of your profile page, at the likely expense of organic interaction with your actual connections, not to mention your attention span.
Points of Differentiation
There exists an alternative. A network of candidates and employers free of recruiters and sales pitches, without tiered-off features and sacrificed monetization. Whereas these features listed below might seem small by themselves, they give confidence that in order to make a connection, one doesn’t have to revert to a shotgun spam approach with a blunt object. A sharper, more personalized outreach (and as a result, better community) is able to occur because of the efficiencies around these controls and the transparency which they offer.
Being able to perform basic boolean searches with AND and NOT operators is a feature that AngelList surfaces through the use of ‘filter pills’. These don’t just help isolate, but also performs subtractions from a result set that could easily become unmanageable if the net is cast too wide. I’m often looking for candidates that are not, for example, “interns” or from “General Assembly”, giving the searcher the ability to filter out candidates that have these terms as a listed component of their profile. Yes, LinkedIn has a global search that will allow me to see results across candidates, companies and user groups – but when recruiting, I’d rather have a more focused search that gives me context as how to reduce down my result set to the best set of candidates.
For any consideration, neither LinkedIn nor AngelList will contain enough information to accurately make a judgement on whether or not the candidate is a fit. The best thing either site’s experience can do is to make it easy for me to access other points of reference, rather than conceal it. AngelList accompanies the messaging that a client and employer are mutually interested with exposed links to the StackOverflow, Github, Blog, PDFs from the email itself. As soon as a viable candidate hits my inbox, the next step for me is to dig further into their background. Not making the employer have to login / re-enter the site, then hunt and peck for details such as those is a great understanding of workflow.
Asking for a designation of status, to whether they’re Just Browsing versus Actively Looking is a subtle leveling of candidates. A concept likely borrowed from dating sites, this acts as a tier of immediacy – something that is especially helpful when searching for those who have a sense of urgency in their job hunt. Just Browsing requires a little more effort to pull that likely already-employed candidate away from their chair at their current job. I tend to skip those because of the exponential effort involved, unless I’m looking for something extremely specific and can’t find any Actives.
When screening candidates, what better tool to have than to identify someone that is currently active in their job search? This single data point alone takes away a great deal of guessing about who’s currently looking for a position, and who is content. Such a simple feature has had a tremendous impact on my response rate of candidates. Two, motivated and actively seeking entities of both employer and employee can now find each other easily. My screening criteria typically starts at 7 days, then expands out to 15, if needed to increase the candidate pool. Last active reduces the amount of dead-ends in the initial Introductory outreach, the task that takes the longest, given the effort of personalization involved.
AngelList exposes to the employer the number of stated interest actions that the candidate has made. A reasonably low number of companies that a candidate has interested in is a more desired quality than a high one. However, the opposite is true for companies – the more candidates interested labels them as a ‘hot’ commodity, and as a result, a more desirable place to work. No idea if one can buy fake AngelList accounts to pay for stated ‘interest’, much like Mitt Romney and his binders full of fake Twitter followers, but I’d love a flood of more / higher quality candidates only interested in the latest hotness. Kidding.
We had a recent interview with a candidate who had indicated their interest in 180+ startups, including our own. His behavior at the interview was consistent with his abusive behavior within the AngelList community. The candidate indeed was all over the place, erratic and didn’t know what he was seeking, career-wise. During the interview, he even inquired about setting up a bot (using Ghost.JS) to automate his interest clicks for him. As a result, I’m paying more attention to these Interest actions taken by candidates as another data point for which to filter upon.
Saved Searches & Alerts
AngelList allows you to save a set of search criterias for easy recall (e.g. – Engineers, San Francisco, Actively Looking, Last Active within 15-days, NOT Intern), which is nothing new, but the ability to automate these search operations as a background process is immensely valuable. This batch job, running occasionally (seems like it triggers every few hours), and sends an email should a relevant candidate create an account, change to ‘Actively Looking’ status, or even login onto the site (making them Active). Not being afraid of losing their own traction due to a lack of return visits, AngelList leverages a basic example of their own machine learning / business intelligence on the users behalf.
Archiving in Sequence
A standard search interface dictates the user to query for a results, wait for a return, then scan down and click to isolate a record. If the found record detail does not match, return to the initial state on the results page and seek a second record. This model is archaic and obsolete, in that the user has already dictated focus on a the record, but the system breaks this context in having to go back to the results page.
However, AngelList’s model of search and browse is built around an interaction model of efficiency. Instead of displaying the entire, full profile in the candidate result set – an abbreviated ‘card’ is shown for quick processing. First acting in the role of screener, I only need to know a finite set of reference points, not see an entire profile. Location, Experience and Skills – all of which takes me less than two seconds to scan and decide to either Archive or Star . Flipping through these in quick succession saves me a great deal of time by retaining my context of focus. LinkedIn has since recognized their initial shortcoming, and recently has patched together a rather cumbersome overlay display to traverse to the next candidate from a single profile.
This may not seem like much in the way of features, but starring gives the ability to ‘cache’ candidates that have matched your screening criteria. This vastly helps the workflow of candidate outreach. Reason being, as you’re screening for one role or the other, the first task is to screen and cache a pool of matches – and only after there’s enough to process (several dozen, in my case), then start to personalize messages of interest. I often aggregate these candidates over the course of the work week, then on Sunday evening when things are slow, complete the task of sending introductions. I know I have a better response rate on Sunday evening when most are dreading going back to their unsatisfying job the next morning. If I were to have to do both the search and aggregation at the same time (as LinkedIn requires) rather than in bulk, it would be a completely different experience.
There are two methods of stating interest to connect candidates with employers on AngelList, but I only use one of them. One can request an introduction to either employer or candidate without having to author a personalized message (a bit like ‘poking’ in Facebook terms), but it seems a recipe to be ignored and best avoided. Much more appropriate is a personal note citing the matching commonalities (e.g. – tech stack, skill set, domain knowledge) and use this personalized outreach as a first impression. It shows you’ve done your homework.
After the outreach between employer and employee has occurred, AngelList will connect the two parties and step out of the way. Contrary to LinkedIn, who wants all communication to happen within the platform itself, AngelList understands email is how people are going to talk to each other anyway, and CC’s both parties with visible email addresses. The AngelList platform has done it’s job, bringing two people together and letting that organic connection take over from there – it is an elegant move to step back (and state as such), rather than try to harness and control all communicate.
Areas of Improvement
As AngelList initially was solely a platform to match investors and startups, there still exists artifacts of this prioritization in the profile pages of all its members. Below the now ubiquitous full-span header image, are tiles of Employer / Investor / Founder references, with no titles and no descriptions. These have a hard time visually standing on their own within the layout of the page. Few members, especially juniors but myself included, have enough work experience to balance this portion of the page properly.
The weight given to adding / changing different Experience tiles is really geared to an Investor who is trying to showcase a whole portfolio of dozens of companies, with their logos tiled in succession, like a VC page. Even the call-to-action to list a new Role as an Investment / Employer / Startup is the first input prompt on the page – but how many (other than serial investors) would use this more than once or twice a year?
The overall scanability of the profile page for a candidate is very cumbersome, with all of the open-form text relegated to the right-hand side. Reading text in that narrow of a column is best left for newspapers, not high-resolution displays. Either I’ve got to be much less verbose in my writing (no doubts there), or get a whole lot more references to balance out the page properly.
I also find it lacking that there are neither descriptions nor titles regarding what role employees had at their various companies. I get this as it relates to work culture – we are all doers, not talkers and titles should be irrelevant. Yet, when trying to grasp the weight, experience and seniority of a client – I need more than just a company name. Yes, this display method is cleaner initially, but I’d like to go further into detail somehow and the Achievements section isn’t the place to do that. This might be another artifact of when, if you founded or invested in a company, that really needed no other description.
Lastly, the separation of Activity into a separate tab for profile pages is an odd choice. It was very the nature of temporal activity that made social networks so compelling in the first place – every time you refreshed the page, there were new lolz from your friends and lovers! But AngelList boxes this in, again relegated in the narrow sidebar in a singular view. This article mentioned some of the killer features around temporal components during the candidate recruitment search, but why this mantra of transparent activity isn’t prioritized on the profile page as well is strange.
How to fix this? Introduce some classification for templates between the roles. Give the profiles of Investors and Founders a bit of different treatment than that of the Employees. As AngelList’s member base widens, there will be an even greater contrast between the big and small fishes in the pond. This needs not to mutually exclusive, as most Investors were Employees too, at one time.
I think the root of the problem would be helped by a better prioritization of the occurrence when user-generated content occurs on a long-term scale (new job, new investment – happens not very often) versus a small one (status updates, new connections – happens all the time). In turn, the profile pages will seem fresher, more balanced and better tailored to the level of AngelList member that it represents.
Extensibility & Integration
AngelList’s API has a history of already being built upon quite successfully, resulting in some great results from independent developers – it is just formalizing this community to integrate something more tangible than just porting data out to visualizations, plug-ins or various widgets. Currently, there’s not anything that can be built into AngelList to help gain a more relevant experience for its members. I’d love to see some of the same type of integrated, external components that Amazon decided to kill early on. Salesforce / Heroku / Marketo are a great examples of SaaS products transitioning from a single site into a full-fledged platform through the enablement of their partner apps.
As a specific example, GitHub’s activity visualization is a wonderful way to discern the activity level of a developer – allow them to embed these components into their profile. Given that this is the first thing I do after being introduced by AngelList anyway, partnering with GitHub and allowing a developer to showcase a visualization of his repository activity would be immensely helpful. I’m sure there are dozens of these examples of industry-specific integrations, not just for developers, but design, investors and other roles as well that can help improve the profile page, for reasons mentioned above.
This article began by describing an environment where the commodity of professional connections is extremely high, advocating that the current industry-leading tool is neither providing a positive experience, nor treating their members very well. The next section illustrates specific points of differentiation between the incumbent networking tool, LinkedIn and the challenger, AngelList. The article concludes with some suggestions on how AngelList can improve their experience, both within the boundaries of their site as well as how they position themselves with content partners to create a more adaptive platform.
The irony is not lost on me that I’ve now become the recruiter whom I initially expressed so much disdain for. However, I’m representing the greater good of our company as well as respecting the nature of organic connections on AngelList. Those aforementioned points of differentiation has allowed me to work more efficiently, enabling me to craft highly-personalized introductory messages and target candidates with unparalleled accuracy. Again, it as a result of these features (and the need to be associated with a company), I advocate that AngelList has created a more open, flat community who will continue to contribute and grow their member base larger and larger. It will be interesting to watch if / how AngelList transcends the niche startup culture into something more widespread, but the foundation of the principles which craft great experiences is solidly in place.