Gagan Saksena

A disturbing trend in usability

2011-11-07T11:22:00.000-08:00

Developing great looking interfaces with good usability is hard. And I certainly understand developing newer interactions that are forever pushing the boundaries of getting more done in ever shrinking display real estate.

One of the things I'm proud of is having been involved in setting accessibility group in motion at Google. It's good to see when Google makes changes that help more people use the web at large and get access to the information they seek. An integral part of such changes is that Google pays a lot of attention to usability through extensive testing. Or at least it used to.

It pains me to see interfaces that are changing for worse. Case in point- the new interface (which will become the default soon) of GMail, Google Docs, Reader, etc. There are number of problematic changes that don't make any sense- For instance, the labels and the chat widgets each have their own scrollbars in GMail. Why is this a problem? It breaks usability by requiring mouse action. Another example- the actual content (in GMail's case- the first message header) is pushed to about 300 pixels out and 282 pixels down! Why is that a problem? The screen real estate is limited (and not everyone uses big fancy monitors) And it's not that the content before that has something meaningful that justifies this change- most of it is white space! This doesn't look anything like a design for web interface. Here, check it out-

White space in GMail

White space in Reader

The real problem- Now I'm not typically known to complain about user interfaces, but the problem that I'm most worried about comes from some casual conversation with my friends currently employed at Google. What surprised me most was that each and every one of my Googler friends that I had a chance to talk about this, didn't like this change! And yet it is live.

There are two possible inferences from this- either Google did conduct a usability study for these changes and found the most unique set of users that love this new UI and rest of us (by my casual survey ~99%) are now SOL. That seems highly improbable. The other possibility is that they didn't conduct a usability study, and as in the history of doing stealthy/creative launches, have a really big plan for all that white space.

I don't know which one it really is the real issue, but either of these options are on a disturbing trend.

If you understand this field better, or know why these changes make sense- go ahead and comment on this post and let the rest of us know. Otherwise help spread the word about this disappointing trend in usability changes.

Series A funding statistics

2011-03-01T21:58:00.000-08:00

I recently spent some time looking at typical Series A investments. I used crunchbase to get the underlying data. And then applied a few filters (like Series A only, USD only, etc.) to get to these results. Here are some important notes with that regard-

The observations are from companies in the crunchbase db.
Only companies with Series A funding in USD were looked at.
Resulting rows: 3730

Here are some interesting observations.

Log of Series A is a normal distribution

Median Series A size: $4M
Mean Series A size: $7.05M

Distribution of Categories

Series A over the years

Series A by Categories

While there are several ways to read into what you see here, it is fun to build your own hypothesis to match the graphs. (For me, I enjoy looking at the outliers that defined 2007/2008) If you believe you have a good hypothesis then please share!

All the analysis was done with the excellent open-source tools R and Python. Email me if you want the scripts or a job pulling fun statistics from large data sets. :-)

Sometimes

2010-11-19T09:47:00.000-08:00

Sometimes I wish I had the time, the patience and the courage to write beautiful words like my friends do. Sometimes I can only admire.

Defining Entrepreneurial Success

2010-05-15T07:53:00.001-07:00

Just how do you define success for an entrepreneur? Is it a round of funding? Is it certain revenue/earnings? Is is purely product milestone based?

The interesting part is an offshoot from another conversation which identifies the joys of entrepreneurship with just being in it. The adrenaline rush for small daily successes, and the little panics that set off when things don't go your way- these make up for the key moments that you live as an entrepreneur.

If the journey is so beautiful, does the destination matter?

Entrepreneurship chat with Raj Raheja

2010-05-14T19:11:00.000-07:00

Sometimes you meet an individual for an half hour and their simple views and insights become so much more valuable than anything else you may experience that day. Today that happened to me- with Raj Raheja, CEO of Heartwood. I know him from a common friend- but then I don't think I ever knew him really. You know, you meet briefly in social settings and have little to learn about what they do at work. But today was different- I was at a conference (which otherwise left me underwhelmed) where I got a chance to talk to him. He had some absolutely brilliant insights from his experience building his company- I wish I could have captured the gems well enough, but then it would take away the opportunity for him to write about it all (which I did encourage him to do!)

In any case, I was reflecting upon how we expect certain things from life, the day, and then something rather unexpected happens when you engage in a conversation which gives everything else a meaning. Today I had that sort of a conversation that made going to that conference worth it. The contents of that conversation will really have to wait till Raj approves of it being published :-) but the core theme was on thinking big; recommendation for and principles from Rob Ryan's Smartups book; and rewarding the right employee that makes an impact- instead of the usual performance crap we've all learned to detest. One thing that stuck well in my mind was how a particular quote inspires Raj- Here is how I remember it- "Earn enough money to afford the time to do the things money can't buy!"

I assure you it was one of the most real entrepreneur stories that matter, the one that inspires you- the one that you can directly apply to your situation.

Wish we could have conferences that just encouraged such real stories and enabled interactions with real entrepreneurs. Thanks for that half hour Raj.

Experiments for Early Adopters

2010-05-09T09:52:00.000-07:00

This has been on my mind for a very long time, but today's article in NY Times inspired me to post this. I have often wondered how if at all would we measure the enthusiasm that early adopters bring. True that in some cases it may very well be just the act of convincing someone to try out your product, but that's not the early adopter I'm talking about. I am thinking more of the ones that willingly step up and grab the next product from, oh... say Apple, for a price that they clearly are very elastic about. In other words, if the iPad was another $100 they will still pay, because price is not such a factor when they are buying it. Or maybe it is! An important side effect of my formal education was that it re-enforced my belief and reliance on quantitative analysis. So, just how could we measure what a new iPad is worth to the early adopters.?

I can think of two interesting experiments. They may sound a bit what Prof. Dan Ariely might want to do- and I encourage the opportunity for anyone to try these out. For the purpose of these experiments I will use iPad as an example of a hot new product.

Experiment 1. Suppose you were given the opportunity to participate in the "early adopters" of iPad with this condition- You will pay exactly the amount that Apple will charge one year from now; you will always have the latest version (thus reducing if not eliminating any price-differentiation or quality issues that an early adopter may face) The only condition for that one year is that you may not use the product in public, tweet/blog about it or in anyway be seen with it. In other words- no one can find out that you own one! (This takes out the extrinsic "fashionable" joy of owning a hot new product) Who will opt for this?
Experiment 2. Exactly like above- opportunity to participate in the "early adopters" of iPad. Except that in this case, it's just the reverse- You are required to tweet/blog about your new product. The only condition here is that the product you will get for one year, won't actually be working (In other words, we have taken out any intrinsic "value" of owning it- except maybe for a paperweight) How would you as a user know what to talk about? Suppose that was available in some form (internal reviews from Apple approved for release for such experiment) So who wants to participate in it?

There are most certainly a lot more details about this experiment that will need to be thought through- like finding a group that's unbiased, understanding the percentage distribution of participants in both experiments. But I hope this serves as a starting point for a thought experiment at least!

I wonder if these figures would correspond to the percentage of early adopters that do participate early on and help iron out the kinks for the rest of the market.

2009- The year of Ubuntu

2009-03-16T11:52:00.000-07:00

I've been patient with Windows. Letting myself boot into Vista every now and then. However off late I've switched primarily to Linux- specifically Ubuntu. As it turns out I'm not the only one making the switch. Mr. Dvorak (of the PCMag fame) also likes Linux.

These are some very interesting times. The play-field is relatively leveled. People are willing to do more to save and they do expect more for their money than ever before. Everything from operating and switching costs to the growing phenomenon of moving your applications, and content (documents, media, etc.) to the "cloud" is going to effect what happens this year. And it is because of these interesting times that I predict that 2009 will be a significant year in establishing Ubuntu as the operating system of choice. Here is my reasoning.

A fresh look at Ubuntu

First let's talk a bit about Ubuntu itself- I'm not going to spend any effort convincing you about how reliable, secure and efficient life has become with Ubuntu, or how far we have come in making Ubuntu a solid platform. This is something you need to experience yourself.

Switching costs
A very important part of reducing the switching cost is the Live CD concept which allows you to try out a full fledged installation of Ubuntu on your hardware without having to install anything. Don't like it (and that would be very unlikely) then you can reboot to your usual. Like it but not sure- you can continue to work with booting from the CD. Like it enough but not sure if you are ready to switch, you can dual-boot your machine to allow you to boot to Ubuntu or your other operating system. Love it? Then a fairly easy installer helps get it up and running on your machine.

The other side of the story is the negative switching costs of moving away from Windows. If you are going to miss your Windows specific software then you might be in for some surprises on that front. One might argue that the familiarity with the operating system is worth staying with it- but there is a fascinating and unintended acceptance of mediocrity that you won't realize till you experience an alternate operating system. Do try other things out and if you still prefer your malware-susceptible, crashing-every-so-often operating system then hey at least you tried.

Operating costs
One of the most impressive features of Ubuntu is its ability to work with fairly modest hardware. This is quite an unusual contrast when you look at the requirements of Vista (or even Windows 7) This is an important factor to consider- you may not have to buy a new machine just to get a better experience. Additionally I've often heard the argument of lack of support/knowledge when running a Linux based system. Trust me- we've come a long ways from the days of help files. There is active support and tons of help online for Ubuntu starters. Plus an important part of this is really in the switching- once you know, you'll know.

Your other choice

I'd like to address this part specifically to the users of Windows. I used to be one (and I still do boot to Windows once in a while) so I do understand where you are coming from (but I honestly want to show you where you could go- without relying on a company's tagline of "where do you want to go today?") I really have tried to be super patient with Vista- trying everything from "smart tips" to improve its reliability, to adding a SD card to improve the bootup time (which BTW didn't change as much; but my shutdown time certainly became over 2 minutes!)

The problem I have with Windows is that at this stage I'm too tired to wait for the OS patches, and the security holes to be fixed. I've been a Windows user for long enough to expect more reliability, less crashes and I want that now- not in the next version or an update. The last straw for me was mostly the incessant hangs (and the option of "downgrading" to Windows XP for an additional cost!)

Pricing was one of my most favorite subjects in MBA. And so it pains me that much more when I read that Windows 7 will be offered in 6 different versions (who was the genius behind this one!?!) The "Starter" version offers the ability to run a whopping 3 applications simultaneously. Yes I'm not kidding. And the other end of that spectrum- a "secure" version since the low-end cheap users can afford not be secure.

But what about a Mac?

Mac is a fine option as well but it's not a cheap option. If you can afford it then paying for the looks may just as well be your cup of tea. And as far as I can tell- there isn't an option to try a Mac out- you either sell your whole hearted commitment to Apple or you don't. I'd have loved to point out other comparisons here but really, Mac is a fine choice if it's affordable.

Evolution of our needs

There is also another important factor- the real requirements of what we expect from an operating system has evolved significantly over the last few years. I remember back in 1997 or so there was an effort by Sun in pushing a network appliance/computer (I forget the exact name) that really was a Mac Airbook (without the sexy looks or the laptop part) I think that idea failed mainly because it was a bit ahead of its time. The network speeds and reliability were just not that commonly available as they are today. And neither were such cool options as Google Apps. Today you can actually get by with just a reliable connectivity. There is increasing effort by several key providers to allow more network storage, and to break free from being tied to a specific hardware.

Challenges

Where we go from here depends on several things- and some of them could present us with some interesting challenges to solve. But I do feel that none of these challenges will take a long time to overcome and so I do believe that this is truly the time when Ubuntu will shine.

Other distros

One of the significant problems with Linux adoption has been the choice of distributions that have been available. A few years back (when the going was all good and the Internet bubble was still growing), RedHat had a real shot at emerging as that leader. In my mind RedHat is now totally off the contenders list. While Fedora is a solid distribution in itself- it is not going to be sufficient in recovering the true Linux market-share. Never before has a distribution caught so much attention as Ubuntu. Here is the proof- Take a quick guess at which color indicates which distro (from amongst Fedora, Ubuntu and Debian) in this chart-

Confirm your answer with the Google trend on this url or by clicking on the chart.

If there ever was a time for the Linux effort at large to really make a difference, this would be it. And if we have to do it right, then the distraction from other distros needs to be minimized. It's important for open source software architects, evangelists to unify their efforts with Ubuntu. Linux has spent several years trying on several faces- this one fits and looks good. Now it's important to stay with it.

Malware catchup

An interesting concern (primarily of the lag in supply and demand) is of the what-if scenarios with malware. While no operating system can completely protect you from a malware, Ubuntu has so far proven to be really secure. It's possible that in the next few years the demand for Ubuntu increases malware development but it's also rather unlikely a platform for that. And that's fundamentally because of it's open architecture. More likely than a malware emerging on Ubuntu, is its fix. Thanks to a rather vigilant and active community of Ubuntu users, application developers and volunteers. The same can not be said about Windows.

Other stuff that could go wrong?
While we all hope that things would turn out for the better, there are some interesting possibilities that might hamper my prediction for Ubuntu. These include- The possibility of Windows 7 being so kick-ass that it does renew the faith of staying with it. The possibility that one of the other Linux distros canabalizes Ubuntu's market share. Or even the possibility that 2009 releases of Ubuntu would suck! Given the looks of what 9.04 is shaping up to be, I'd have to bet that it's unlikely.

I still hold that 2009 would be the year of Ubuntu.

The Marketing Farce

2008-10-30T10:34:00.000-07:00

Before I started the MBA program I had an established bias of how all marketing was a load of crap... Wharton's classes changed a whole lot of that perspective for me. For starters- the whole analytical aspect of marketing caught my attention, and as did the behavioral sciences element of it. I saw that if you put aside the bias, there are fascinating principles, observations and statistics that challenge our current understanding of how well we know and understand the human mind- as a participant, as a consumer and as our customer. There is much work to be done ahead- specially as we start factoring the social interactions into our models. (Something that my current company is exploring)

And then I attend Social Media Strategies Conference. And this is where I circle back to why marketing gets the bad name. A "marketing expert" speaker here claims "over 15 years of online marketing experience"- now that puts him pre-1993 days. Wow. No wonder people like me develop opinions like mine about marketing :-/

Maybe Wharton just made me pay more attention to numbers then ever before. Gee thanks! :-)

Lessons from former Googlers

2008-10-11T11:00:00.000-07:00

I am on a group of ex-Googlers and some recent threads over there caught my attention. I realized that there were some key lessons learned that could be useful to Google and other companies in general- lessons on hiring practices, managing people, compensation and incentives, etc. These lessons also apply to some degree for candidates interested in working at Google (or other similar companies)

Before I dive into these observations I have to say that I left Google for my own reasons and it was by far the finest companies that I've ever worked at. It is a great company that I absolutely respect and admire. I really am thankful for all the great experiences I had at Google- From playing volleyball with Sergey to having crazy MySQL optimization discussions with the amazing Stacey. My pleasant experiences certainly outweigh the collateral annoyances expressed in the summary of fellow ex-Googlers below. But just like every company I've worked for, they are not perfect- the lessons here help us understand where they could do better (and for any company that can learn from these observations)

The hiring process

Confusions along the way- Google has a rather notoriously painful hiring process. One of the common confusions in that process is to get interviewed for the wrong position. Interestingly this observation was also noted from not just the ex-Googlers but also from some people that I had referred. Word of advice to candidates ready for their interviews at Google- always double-check the position you applied for with each person interviewing you. Sometimes this small up-front check may prevent a rather unfortunate result of a mismatched interview. If you are a company in this situation- keep the overall process transparent and as quick as possible; pay attention to the position and in case of a discrepancy, make it clear to the candidate why he or she is being evaluated for another position than their expectation.
Low offers- The typical offers from Google are low. We've heard- "People don't work at Google for the money. They work at Google because they want to change the world!" The real observation here is that Google makes a concession for its brand name on your resume. Personally I think it's the market efficiency of brand names. Candidates- just factor this in your expectations. Companies- don't appear cocky about your brand- remember that this is a dangerous game to play and for most candidates a good offer is fundamentally its monetary value. You could get by with lower offers but this will in turn effect your retention rate. Once a candidate is done with the brand name on their resume, they'd be more than ready to seek a more fulfilling offer elsewhere.
Long process- The hiring process at Google is long (typically 90 days or more!) Sometimes this has made it difficult for a candidate to wait that long. As a candidate you need to be prepared. And it's exceedingly challenging to weigh your offer with another unless you time it well. If you are a company- keep in mind that with a long process you may be losing a whole set of great candidates (specially if you have already spent time and effort in filtering them through some of the early levels)
GPA and keyword cutoffs- Google used to receive a large number of resumes everyday when I was there, I suspect it is still very true. If a company has to go through a large set of resumes it has to resolve to a more scalable process like matching for keywords in resumes, or set an arbitrary cutoff like good academic pedigree, or certain GPA levels. If you are a candidate- you really need to work hard at building a very specific resume to the job that you are interested in- following very exact hints from the job requirements. Leave your GPA out of the resume if it is not spectacular- let a terrific experience build your case. If you are company facing this problem (and I don't know very many in today's world) then build a better filtering mechanism that does weigh in other factors and not just do a keyword match. Understand, for example, that there may very well be some excellent engineers that almost feel beneath their level to claim familiarity with hackneyed resume terms of "object-oriented concepts" when they have designed libraries for core Java.
Little to no negotiations- An interestingly common observation seemed that there is a "take it or leave it" attitude when dealing with Google offers- mostly resulting in candidates taking it with relatively lower happiness. If you are a candidate- keep that in mind as you get closer to your offer (spend some time learning simple negotiation techniques) and ALWAYS negotiate even in spite of what this observation tells you. If you are a company, and feel proud of your negotiation skills, then keep an extra layer of sweetener to offer once the negotiations are done. This will ensure that the candidate joins your company with the level of enthusiasm and content that should come from a "happy" offer.
Little transparency on job details- Back in the earlier days of Google, it was very common to be expected to join without knowing any specifics of your actual job. I personally know of at least one great candidate that refused an offer because Google wouldn't tell him what he'll be working on. I strongly suspect that this is now changing, but if you do experience this still here is my tip for candidates- Ask. In most cases you will likely be interviewed by your prospective peers and/or your manager. While a direct question like "what would I be working on?" may not get you much of an answer, an indirect question from which you can build a good inference could be more beneficial- "where do you see me as a good fit in what you do?" If you are company, do whatever NDA signing you have to do but make it more transparent for the candidate to get a better picture as he or she gets closer to the offer stage.
The line outside the club- There is a fascinating marketing effect of the "line outside the club" that subconsciously makes us believe that what's on the inside must be worth it. If you are a candidate and know of such a line, don't jump to any conclusions. Evaluate what it would mean for you to be working at such a company (and the relevance of a specific job requirement) and your ability and willingness to wait in line. If you are a company doing this, be aware that you may just as well be turning away the really good candidates that are rushed by their circumstances.
The recruiter- The typical recruiter is a contractor- with a term limited to 6 months (I don't know if this is industry specific) Given that the average process takes about 3 months, you can see that you will deal with more than one recruiter 50% of the time. Which means you as a candidate, have a coin-toss probability on either dealing with the same recruiter throughout your hiring process or be dealing with a newbie who is also juggling with taking over the list of prospective candidates from another/previous recruiter. Rather common and surprising expression that's often heard- "I think you are in, but bear with me as I learn more about Google."

An unfortunate result of having short-term contractors is that they never really get up to speed on a certain job family well enough to make a positive difference. While it's easy for them to do academic pedigree checks or GPA cutoffs or keyword matching, they don't make job specific logical deductions. Eg. if the job requirement says the candidate should possess skill A and the resume reflects a job B which implicitly means that skill A was used but the resume does not specify skill A, this resume is most likely going to get tossed. The real sad part here is that most recruiters will typically have very little dis-incentive for making such mistakes. The large volume of resumes (and the fact they are over-tasked) makes it very easy to move on to the next case. As a candidate, you should really work hard at stating the obvious in your resume (without of course going overboard with it) The best way to get that is to literally have someone else review your resume.

And to address the changing recruiter issue, if you are a candidate in this long process, build a good relationship with each recruiter but don't fall in love. They are professionals trained to give you the warm-comfy feeling about this process but they will switch their jobs on you. If you really want to factor this in your process, then ask your recruiter about how long they have been at their current job up-front. If you are a company and under the typical 6-month restriction then try and reduce the hiring process time or get rid of that 6-month resitriction, or hire recruiters! Keep in mind that the recruiter is the face of the company and changing faces in the middle of the process does not give a confident appeal to candidates. If you really can't improve on this, then at least set recruiters by appropriate stages of the hiring process so that a more natural and logical hand-off occurs when say a candidates goes from phone interviews to on-site.

Life as a Googler

The perks- While there are numerous perks at Google that make it the "best company to work for" most of these perks are, in the words of an ex-Googler "family unfriendly." If a company offers free dinner it indirectly discourages spending enough time with one's family. This of course is less likely the case, if you are 20-something or are fresh out of college and love the campus feel of hanging around when you should really be socializing outside your work. If you are such an employee that's wide-eyed about these perks, take a step back and evaluate your overall situation. You are likely going to be very disappointed the day you decide to leave a company like Google and face the real world. If you are a company with such perks, be careful. You don't just want good dedicated employees, you also want happy ones that do have a life outside your campus.

Another observation in the perks department- there is a serious dis-balance in the utility of a perk being offered and the same perk being taken away. Yes this is a fascinating example of mental accounting and Prospect theory at work here. To illustrate this take a look at this graph.

The real interesting problem is when this starts to compound. Multiple adjustments do get compounded and the combined effect is much more pronounced. There is also an interesting thinking at play here which says "if I have a justification for why my free soda was taken away and that same justification seems to explain why my free massage was taken away, then no matter how ridiculous it sounds, my justification must be right."

There isn't much to do as a consumer in this economic model of perks- you will mostly do what your personal behavioral finance guides to you to. But there is a lot to be learned if you are a company thinking about offering perks to your employees. Don't offer what you can't sustain.
The actual work environment- The actual work environment is way more competitive than you've ever felt at any other company. A common observation amongst several ex-Googler friends of mine- it can be very demoralizing if you aren't the best in a group. My advice- If you are not the best in the group, don't take that to heart. Your contributions to the success of the group and the company are still going to be valuable whether you won the "best employee" award or not. If you are a company in such a spot, you have much more work cut out for you! While you want to encourage healthy personal and company growth, keep in mind that too much competition within the same group can be disastrous. Good and effective teams should have several "best employees" in non-conflicting categories. Awards should hence be focussed more on teams than individuals.
The attitude- There seemed to be certain cases of discrepancies in individuals' behaviour in a team based on certain distinctions. The common ones seem to be whether you were an early employee or not, or whether you are an engineer or not. These are serious threats to the operational efficiencies of a team. These unfortunately were not uncommon observations. You do need to lookout for such problems and address them early on before it becomes a serious issue. This applies to both individuals and companies.
The people manager- Most of the observations were on how poorly the relationship developed (or didn't) between individuals and managers. At least for the initial part of Google's explosive growth, the ratios of managers to direct reports were very high. While this supposedly kept the organization structure pretty flat, it didn't help build the much needed relationship between an employee and his manager who has 80 other reports. Companies need to take note that managing people is serious business and not just another task that a "technical leader" should grow into.

I do believe that these issues are all addressable to really make a company like Google, as perfect as it can get. If you were in Google management and had a say in it, what would you have done to retain some of your best talent?

Resuming the blogging life...

2008-05-03T11:25:00.002-07:00

It's been a while. Since I wrote anything worth putting on a blog. But I really need to do this. Not for anyone else but for me to get back in the writing mode. School's over and this is as good a time as any to practice my writing. If you happen to be the unlucky reader of this blog, send me your thoughts- on how I can improve my writing, on what you disagree with, on what you thought was interesting. Or just ignore my little exercises in writing :-D