Brendan Baker

Keep ratings ratings, and discovery discovery

I love Netflix in many ways. 

But this is frustrating. This is taking a community/market norm - star ratings - and applying it to recommendations.

I want to see what the world thinks. I don’t want to see Netflix’s Brendan filter on what the thinks the world thinks. 

Beyond the frustration of just trying to figure out if a film is going to be good, this is a symptom of a darker trend of personalization technology increasingly isolating us from public consensus and opinion. 

Your buzzwords aren’t fooling anybody

“…may have an interest in our very innovative seed-stage company, XX, which is a disruptive technology in the e/social commerce world. Our technology will cause a paradigm shift in the multi-billion dollar ecommerce XX industry as it allows the customization of XX and XX and same day delivery.”

Just say “we let anyone create and customize XX, and have it delivered the same day.”

Just say what you do. 

Your buzzwords aren’t fooling anybody.

B

(actual pitch received; all sector/startup identifying elements removed)

Your Space is Crap

(Statilizer screen cap, from their website)

Social media engagement and analytics:

SAS SproutSocial ArgyleSocial SocialReport Campalyst Unilyzer Curelate Pinfluencer SimplyMeasured socialbakers PostRocket Adaptly Reachli LittleBird Sparked Shift Traackr FanBridge HootSuite Morcova prosodic Swix Venueseen Monstro Twimbow Buzznumbers TrendKite Sociero Youscan TunkRank Echnobot SocialAppsHQ Vaktarrin Beevolve Mention Buzzware GovSM Casapplanca SocialLogix Morcova ZocialInc BrandFractal ejenio Croakun StatFly SocialKnow position2 Hubspot minilytics tweetreach socialoomph sodash Decooda Teezir motivequest SocialFactz Kantar iMonitoring Meltwater Engagor infoFactory NetBreeze NutshellMail BoomSonar ReviewIQ AtInternet Daumsoft Infospeed Wool.Labs Actionly SocialPointer Zelist Telligent GoogleAnalytics ReputationObserver Wildfire TodayLaunch ethority ChatterBeacon Katapedia Blogmeter ReputationTool Webfluenz Scup Netbase Finchline Trendiction Digitalmr Brandreact Mutualmind Semioboard ebuzzconnect Memonews Internet911 Informant.se twentyfeet Social360 Replise Geofeedia Looxi SocialTalk Clipit Spredfast Rightnow Topsy Imooty Tealium EarlyDetection Mindlab.de TipTop Tracebuzz Brandtology Thoughbuzz JamiQ Insttstant Tweetlytics Overdrive Mentionapp Repumetrix Socialscape eWatch Cision Digimind Samepoint Buzzstream Vocus Attensity Buzzlogic Factiva eCairn Jive Mediamiser Moreover Onalytica Reputation.com StartPR.com UberVu Lithium VisibleTechnologies Spiral16 CollectiveIntellect NetworkedInsights Backtype nminsight Trendrr SocialMention Monittor linkInfluence nms CrimsonHexagon VisibleMeasures CustomScoop Attentio Tweetbeep Collecta Tweetfeel SocialRadar Synthesio YackTrack Brandwatch Addictomatic Echometrix SentimentMetrics ExactTarget HowSociable MediaVantage SearchMonitor Kana Postling PeopleBrowsr Twitalyzer EdgerankChecker Vrank Brandify PageLever Crowdbooster Tweeb SiteTrail Kyoo Sysomos BuddyMedia 

It goes on. 

Of course these aren’t all identical. That’s not the point. The point is that it’s a very crowded space. there are hundreds of products that…kinda look the same, even when they’re not. 

When I reviewed 8000 pitches at AngelList, I saw waves of similar startups, again and again. Every well known space has dozens of competitors, at least. Not only this, but while the investors see many of the players, founders only know of a few - theirs, a couple they’ve stumbled across, and the two high flyers they’ve read about on Techcrunch. So they’re at a disadvantage in knowing how to differentiate. 

What does it mean when you’re in a crowded space, or a space that has seen a long line of failed attempts?

Do not define yourself by your space. You’ll be too easy to dismiss by fatigued investors and customers, who can’t figure out what makes you different. 

Do not define yourself by your competitors. Learn from them, but focus on your customers. The battle is the for the customer, not against lemming competitors. 

Define yourself by what you’re actually doing. Tell people exactly what you’re building. 

B

Zombie VCs: Using Numbers Effectively in Silicon Valley

Danielle Morrill had a characteristically aggressive post yesterday about Zombie VCs - those who are slowly dying and can’t or won’t invest in your Series A. The recommendation was to ignore them. Reaction ensued. 

This is an area I care a lot about: transparency in our industry, accountability, better feedback loops, more efficiency for founders, and, above all, timely and relevant quantitative-based analysis. So props to her for the effort. 

Here’s my take:

1) There’s too little numbers-based analysis in our industry. What we do have is high level economic analysis - not actionable. Danielle took a strong opinion, found data that applied (more on that in a bit), and used it to support a conclusion. Good for her. That is a positive effort. This should be much more common. 

2) There’s little doubt that the feedback loop here (VC fund death) is too long. In fact, the feedback loops in our entire industry are far too long. This costs us in wasted resources. Acknowledging failure and learning from it is healthy, on both sides of the table.

3) Efficiency in fundraising is key. It’s partly a numbers game. Few founders deeply understand this, so this is helpful. Founders should prioritize ruthlessly on how likely investors are to invest. Investor activity is one factor. 

4) Strong stances are good. Mostly. Danielle’s approach is always an assertive one. Generally I like that. Our industry is filled with soft, anecdote-laden, non reputation threatening, general advice and guidance. Many fewer people call bullshit so assertively. We need more bullshit-calling.

That said, her credibility suffered here when she didn’t leave herself a little more room on the uncertainty of the data. Data in our industry is sparse. We must use it for guidance, but we also must acknowledge the ways it can be wrong. Saying ‘here’s my conclusion, take it as guidance but understand that it may be slightly wrong if X or Y is the case’ would have been a good idea.

5) The data approach here isn’t that bad, overall, but there are important cases where it’s wrong. 

Some (many?) in the zombie list are legitimately active, or trying to be.

- They may not have seen deals that didn’t fit with their criteria (dry spells happen, even to great, active VCs).

- They may have tried to invest, and seen their term sheet beaten out by another investor.

- They may have invested, but those investments are still private. Common, especially in enterprise.

- They may have invested and announced it publicly, but that didn’t make it into the data set. Common. 

- A 6 month analysis window is too short. A typical VC might be involved with 8 companies, who are each active between 4 and 8 years. Do the math - that makes for fairly infrequent investments. 12 months is more realistic, for such a decisive conclusion. 

All of the above contribute to a gap between the set of investors who are legitimately interested and able to invest right now, and the data of who has reportedly done deals. 

It’s not that this data can be easily found. Data in our industry is imperfect. It’s that when the gaps above are acknowledged as an important piece of such an analysis, it gives the whole exercise more credibility. The ‘screw you, if I’m wrong, then it’s because you’ve failed to report properly’ approach loses. Acknowledge the weaknesses in the analysis, and more people will believe the underlying thesis. 

I thank Danielle for her aggressive analysis and recommendations. I hope she keeps doing this; it’s healthy for our whole ecosystem. But by dialling it back 10%, and acknowledging that the data and analysis methodologies in our industry are imperfect, she’ll be even more effective. 

B

The Perfect Analyst

…operates at an astonishing range of levels.

• Question how our industry works with relentless curiosity.
• Be able to form hypotheses about how things could work better.
• Identify the right data to answer questions. 
• Hack or build solutions to get this data, answer questions, and trigger action from others. 
• Present solutions and support our various teams efficiently, with insight. 
• Deliver with dependability.
• Understand what makes great companies great, and poor companies poor. Have opinions. But always try to understand this better.
• See the deep themes running through our world, from the industries we hope to help disrupt through our companies, to our own.

Most teams would be happy with someone who can work at a couple of these levels. That’s fine if you’re filing TPS reports. 

The perfect Analyst on our team has a bunch of these levels nailed, and will be eagerly trying to get the others down. 

By the way, we’re hiring.

B

Two reasons why the criticism of Mayer’s work from home decision is bullshit: a double standard and the lack of relevant perspectives

I’ve been surprised at the strong and negative reaction to Marissa Mayer’s recent decision to compel Yahoos to stop working from home. Unfortunately it’s largely bullshit. 

A Double Standard

First, she’s being held to a double standard: that she is making it harder for women to work. I’d be shocked if a male CEO was judged on this, as widely as she has been. That’s not right.

She’s not a woman CEO, she’s a CEO. Her job is not to be a woman CEO, it’s to be a CEO. People shouldn’t treat her like a woman CEO, they should treat her like a CEO. (1)

Let’s Use the Relevant Lens

Second, it’s not about remote work, it’s about turning around a company.

Nearly all of the criticisms miss that lens and treat this as a statement on remote work. Nearly all of them miss the difficulty, and extreme nature of turning a massive company around. 

It’s about re-establishing an effective culture. It’s about rallying the great people and cutting away the not-so-great ones. It’s about drawing new talent in that wants to be in a challenging, intense context. The current mix of people at Yahoo will have to change, in effectiveness or mix, for it to become successful again. 

I think remote work is healthy, in the right situations and managed well. At AngelList I worked from Oxford, Istanbul, DC, Vancouver and San Francisco. It didn’t matter. At Greylock I work from home sometimes, and have recently worked from Vancouver and Singapore. 

But those are different contexts than Yahoo, with the team, systems, and culture to make it work. Working remotely is not a worker right - it’s an arrangement that can work well for both sides, in the right context. 

And pretty much all of the work-from-home criticism seems to come from people who are viewing it through their personal lenses. The credible lens is through the management teams of large, recently turned-around companies. If any of those folks want to jump in, I’d love to hear their thoughts on whether this is a constructive part of a larger turnaround process. 

B

1) FWIW, I do feel strongly about bias in tech. I just don’t feel that holding leaders to different standards based on their gender gets us to a better place.

2) Could it have been handled better? Possibly. But that’s not the core question.