elan dekel

Mar 25

Facebook Pages are a bad investment for small businesses

Read my post on Forbes

Dec 18

Shipping Products From Your New Online Store

Read my post on Forbes

Nov 28

Why did Apple fire the Maps product manager?

Read my post on Forbes.

Nov 25

How we set up our online store, Part 3: Domain, Email & Phone

(See Part 1Part 2)

Now that we had picked our ecommerce platform, we needed to get a domain name and email accounts. I also decided to set up a company phone number as I remembered seeing research that ecommerce sites that prominently display a phone number convert better.

Initially we had wanted to pick “lulubee” as the company name, but as “lulubee.com” was already taken, we went with “lullubee.com”, which was pretty close and had a nice ring to it, similar to “lullaby”. I have been using Godaddy.com for several years, and although I dislike how their site tries to constantly push you to buy stuff you don’t need, I have been quite pleased with their customer service. I can always get someone on the phone and they are always technical enough to be able to provide useful help. I bought the domain and changed the DNS record to point to our new Shopify store. 

Next step was to go to Google Apps, and set up an account. This provided us with @lullubee.com email addresses, Google Spreadsheets, Google Docs, Google Groups, and a bunch of other Google services. I made sure to set up a free account which allows up to 10 email addresses. I did some more DNS record modifications on Godaddy, and our lullubee email addresses started working immediately.

From within the Google Apps dashboard I set up a Google Group for our customer support emails, and gave it the support@lullubee.com email address. I set the group permissions to allow anyone within the lullubee team to view the messages sent to the group, but I also selected the checkbox to “Allow anyone on the internet to post messages”. I then subscribed all of the lullubee team to this group. This allows anyone to post a message to support@lullubee.com, but only lullubee team members will see the messages or can access the group. Now whenever someone emails support@lullubee.com, that email gets forwarded to anyone within lullubee that has subscribed to it, and we can go to the group to see all the past email threads. Once we get significant customer support volume, we will probably want to upgrade to a dedicated CRM system, but for now this works fine.

The final step was to set up a company phone number. I didn’t want to put our personal cellphone or home phone number on the website, for obvious reasons, and we don’t even own a landline (who does these days?), so I looked for a virtual phone solution. At first I hoped Google Voice would work, but I was dismayed to find out that Google Voice did not allow multiple numbers to point to the same phone, and as we already use Google Voice on all our numbers we needed to find another solution. I checked out a bunch of them, RingCentral, Grasshopper, Phone.com, and some others, and ended up going with Line2 as they had the cheapest monthly plan, at $9.99/month. I was surprised that I couldn’t find a free virtual phone service. As I was typing up this post, I went back to the Line2 site and it seems that they have added a free package, which I plan to test in the near future. Line2 has an iPhone App that initially gets the call, presumably using the data connection, and it has a different ring, so we know if its a customer calling the lullubee number from the website, and we can answer the phone appropriately. If you don’t answer via the iPhone app, it forwards the call to your regular phone number.

So now we had our basic website up and running on the lullubee.com domain, email addresses for the team, a support email address, and a company phone number proudly displayed on the site.

Nov 24

How we set up our online store, Part 2: Picking an eCommerce Platform

(See Part 1)

The first thing we wanted to do was get a nice looking store up and running so we could put some products online, and start to get real world feedback. The store had to look professional, because we wanted to gauge whether people liked the products or not. We did not want people leaving because the store looked cheap or fishy looking.


My criteria for selecting an ecommerce platform were:

And most importantly:

After some research online, and asking friends, I narrowed it down to Shopify and Magento. Magento is an open source ecommerce platform which was bought in 2011 by eBay, and has a wide range of solutions from a low-end hosted solution, to solutions for large companies. Shopify is a Canadian startup with a fully-hosted solution, at various price points. I signed up for the 30 day free trials on both MagentoGo (Magento’s hosted solution) and on Shopify.

Both seemed to have a vibrant ecosystem of 3rd parties creating plug-ins, providing services, and designing themes. Both seemed to allow integration of many payment providers. Both seemed to have sophisticated admin consoles allowing you to easily add products.

My first test was to see if I could find a nice theme. I was willing to pay a reasonable amount, up to $200, for a nice theme, figuring that this was much much cheaper that hiring a designer. After spending some time on both platforms, I found that Shopify had a very nice selection of high quality themes, and most importantly we were able to find a theme that suited our concept for $150. MagentoGo’s themes didn’t have much variation, and the overall level of quality and finish seemed much lower. Shopify seemed like the Tumblr of eCommerce, while Magento was like Blogger. If you selected Magento you needed to have the budget to hire a designer to either design a new theme from scratch or to customize an existing theme.

At this stage, given that our goal was to get a professional looking site up and running quickly, we went with Shopify. I selected the Basic, $29/month, plan. It only took us a month until we needed to upgrade to the $59/month option, but more on that in another post.

Checking out both options only took us a couple of hours, and after buying a domain name at GoDaddy and pointing it to our new Shopify store, we were up and running that evening.

Nov 23

How we set up our online store, part 1

Over the past six months I’ve been helping my lovely and talented wife, and her business partner Laura Burch, with their new online business, lullubee. Their company designs and sells kits for creating all kinds of crafts, from jewelry to costumes to toys and more. In order to help out, and as an opportunity to learn about ecommerce, I volunteered to handle the technology side of the company. 

Over the next series of posts I’ll discuss the various decisions that we made to set up and market the store, and hopefully help others that are considering going into business online. 

Here are some of the many decisions we had to make, in chronological order, each of which is deserving of its own blog post:

My philosophy going into this project, as it is a new business in a relatively new product category, was for this to be a lean startup as much as possible. We would try to put up a simple site, create several kits, see if we could get customers, get feedback and iterate. I have always been impressed by the now classic tale of how Bill Gross at Idealabs tested the Carsdirect model, and thought that we should do the same:

We put a site up on a Wednesday night; by Thursday morning, we had four orders. We quickly shut the site down (we’d have to buy four cars at retail and deliver them to these four customers at a loss) but proved the thesis. Only then did we start building the real site and company.

Read more: Business Insider

As we were going to be testing the business and the product while setting up the company, this meant that we would have to be able to experiment and iterate quickly, that we would want to take advantage of existing platforms and technologies as much as possible, to do things in a cost effective manner, and to not overbuild, as there is no point in building a powerful, scalable platform before you know your business.

In my next post, I’ll talk about how we picked our e-commerce platform.

Jan 06

An AppStore for Health - Electronic Health Records for the iPhone Generation

The current state of affairs

There have been lots of failed attempts at getting consumers to use electronic health records. Google Health was probably the most visible one, but there have been many many others. Many of the early online health sites created their own online electronic health record systems, in the hope that people would use them to store their health data and then be “locked-in” to that site.

The problem with those solutions is that they all expected you to manually enter in your blood test results, upload your X-ray’s, enter in your glucose levels, your blood pressure measurements, and more. And the only benefit they provided was to store this data for you, in case you needed it in the future. And perhaps they would show you some graphs of data over time. 

These are examples of products requiring a high effort for low reward.

The reason they needed you to enter in the data yourself is because your health data exists in thousands of separate, siloed, computer systems that are not interconnected. Hospitals are not connected to your doctors clinic which is not connected to the lab or to the imaging center. Even within individual hospitals the computer systems are often not interconnected.

In the USA, HMO’s like Kaiser are making progress at integrating the hospitals and clinics and labs within their umbrella, but that doesn’t help everyone else, and it doesn’t even help people that are transferring into or out of that HMO.

What people really want

A much better solution would be a system that was integrated with all of your healthcare providers, and automatically sucked in the data without you having to do anything. Your X-rays would be uploaded from the Imaging center, the blood tests results would come from the lab, etc.

This would solve the pain of having to manually enter in your health data, but it still only solves part of the problem.

What you still need is a way to make sense of the data. For that, the solution is to create secure API’s that would allow app developers to write software that could analyze your personal data, and then do interesting things with it. Correlate your lab results with your Fitbit, with your weight, with your diet. Compare them with publicly available data. Find insights from your 23andme results. And much much more. 

What we really need is a health app store, where we could purchase apps that crunch through our personal data and provide us with actionable information.


The only way to get large, entrenched organizations like hospitals and insurance companies to open up their databases (even to the patients themselves!) is by government regulation. There are simply too many entrenched interests and fiefdoms within these organizations, that getting them to voluntarily open up their databases would be next to impossible. 

The government needs to pull together a group of people that know how to build this kind of thing, and have them specify a system for exchanging the data securely. Then it needs to require all healthcare organizations that hold patients data to make it available to any patient that wants access to it. Steps in this direction are being made, but there is still a long long way to go.

The apps, and the appstore? Thats the easy part. Once the data is accessible, entrepreneurs and investors will take care of it. This is a huge, multi-billion dollar opportunity, and there will be no lack of takers. 

Nov 14

The International Opportunity

One question that people keep asking me is how did I decide to start Medico.com. Not why, but rather, how, did I come up with the idea. On the one hand it seems pretty obvious. Of course every country should have a high quality online health site, but on the other hand its one of those ideas thats almost too obvious, and that one assumes already exists in every country.

There are a number of fairly simple techniques to get a sense of whether there is indeed an international opportunity for your company or idea. I’ll walk you through some of them in this post.

The World is a Big Place

Most companies in the US, and in fact in most countries, don’t think much about the international opportunity for their business. However, the world is a big place, and a surprising number of internet users are outside the USA. In fact, US internet users make up only 11.6% of the internet users in the world.

In addition, one assumes that there are certain inalienable rights for all internet users: search that works and is always available, free web-based email, free voice and video chat, free classifieds, social networking, and many more. Indeed most of these products are available internationally and have significant usage worldwide. This is especially the case for communication and social tools, which users use to create their own content in their own language, even if the UI hasn’t been localized. Examples of these are Skype, Facebook, and Gmail. Facebook had significant international usage even before they released their crowd-sourced Facebook Translations app at the end of 2007, which allowed them to launch Spanish and French UI’s in April 2008. But when you look at other types of online products, you will see that there may be significant differences from market to market.

Consider these two heatmaps, the first for Craigslist, and the second for Facebook.

Searches for Craigslist around the world (normalized by overall searches per country):

Searches for Facebook around the world (normalized by overall searches per country):

Note: these heatmaps, which were generated by Google Insights for Search, don’t compare absolute volumes, rather they compare the ratio of searches for a specific query in a country to all searches in a country.

The only countries in which there is significant interest in Craigslist are the US and Canada. Conversely Facebook is popular in many countries, in fact the country with the most interest in Facebook is Turkey, closely followed by Venezuela, Italy, and a whole bunch of other countries. Of Facebook’s 700M users, only 150M are in the USA. There may be many reasons for this difference, not the least of which is that Craigslist, being a marketplace, is possibly fairly challenging to launch in a new market.

From these heatmaps, one might come to the conclusion that there is an opportunity in the international classifieds space. However, before diving in and working on your Y Combinator application, you may want to stop and verify your hypothesis with another nifty tool. Google Ad Planner allows you to “Search by Audience”, and to display all the websites in a country, neatly organized by vertical and sorted by traffic. Here are the top classified sites in the UK:

From this list, its apparent that the classified space in the UK is quite healthy. Gumtree.com has 10.2% reach, and generates approximately 66 PV’s/user/month. This isn’t quite as good as Craigslists 22% reach and 380 PV’s/user/month in the US, but it is nonetheless a very strong incumbent. Ad Planner is a very powerful tool for learning about the online landscape in international markets.

In 2006 there were many markets around the world without a strong incumbent, and in fact, Fabrice Grinda, even without Ad Planner (it launched in 2008) came to the conclusion that there was an international opportunity in the classified space and founded OLX.com which went on to do quite well

Smart Local Entrepreneurs

In some countries, notably Germany and China, there are local entrepreneurs and investors that specialize in rapid copying of successful US online businesses. StudiVZ (the German Facebook), Qype the German (now European) Yelp, and Youku and Tudou (Chinese Youtube clones) are some of the more well known examples, but there are many more. US companies that don’t think about their international strategy at an early stage run the risk of losing these significant international markets to smart and aggressive local entrepreneurs. 

The International Rollup

Some products require a significant amount of local content for them to be useful, like Yelp for example. And launching a product like Yelp internationally isn’t that easy. In order to launch Yelp in a country, you first need to load it up with a list of all the restaurants and local businesses in that country. This requires a business development person from Yelp to find a local provider in each country that they want to launch in, from which they could license this list, on reasonable financial terms. In fact they probably need to license multiple data sources from multiple providers, one for businesses, and another for restaurants, and another for utilities (post offices, train stations, etc.), and probably many more. And then their engineers would need to learn how to parse the different address formats so they would be able to handle them correctly. And then they’d need to launch it locally, perhaps city by city as they did so well in the US

At this point, the management team at Yelp might understandably have thrown up its hands and said, “lets punt on ROW for now, and lets revisit after the IPO, at which point we’ll buy the leading local Yelp clone in each country.” This, in fact, is a time-honored tradition executed to varying degrees of success by companies such as eBay and currently by Groupon. In Yelps case, it appears as though they are actually trying to go international on their own. Lets hope they aren’t too late, Qype launched in Europe barely 18 months after Yelp launched in the US. BTW - we can track Groupons progress using Insights for Search:


While I was at Google, we analyzed the development of the internet market in Africa. We created a matrix, with countries down one axis, and internet verticals (classifieds, online dating, etc.) along the other axis. We colored in the boxes with green for verticals that were well covered, yellow for verticals that weren’t well covered, and red for verticals that were essentially non-existent. It was surprising how many yellows and reds there were. An enterprising, tech-savvy, computer science graduate should immediately realize that this means opportunity. And they’d be right. Why are these opportunities just lying there for the taking? Again, when you stop to think about it, it makes sense. 

Startup Ecosystem

To create an environment conducive to entrepreneurship, especially online entrepreneurship, you need several elements to be in place. You need a good technical university that creates lots of smart computer sciences graduates. You need investors with experience in investing in startups. You need a legal and regulatory regime that is supportive of startup companies. You need mentors and role models who have done it before. There are very few places in the world with all of these elements in place. And even when all these elements are in place, you need to think about the size of the market.


The US is unique in that it not only has all of the elements in place to create a vibrant ecosystem, but it also happens to be a really really big country. The US is so big, that entrepreneurs can create hugely successful businesses not just in top-level internet verticals, like social networking, classifieds, online dating, online financial services, but in sub-verticals and even sub-sub-verticals. Not only is the market big enough to support all of these niche businesses, but there are enough experienced engineers, investors, mentors, marketers, designers, and sales people to build and run all of these businesses.

In most countries, not only is the market too small to support a niche internet business, but there aren’t enough trained people to run all of the niche businesses that are successfully operating in the USA. Most markets can barely manage to create local businesses to cover the top level internet categories, and many of them don’t have particularly great technology, design, or content. Not to mention the sub and sub-sub verticals. 

Go International, young entrepreneur!

All of this means that there are significant opportunities out there. Instead of launching yet another iPhone photo-sharing app, take a look around the globe. I believe that there are opportunities to build significant businesses both in large individual markets, and also to build businesses that are designed from the outset to be international. For many web businesses, you don’t even need a local presence until you need to establish a sales team on the ground. Facebook is hugely popular in countries in which a Facebook employee has never set foot. One can go a long way using LinkedIn and Skype to meet partners, AdWords and Facebook to get traffic, oDesk and Paypal to hire translators and local contractors.

This is why I founded Medico.com. We think there is an opportunity to build a global health information and community site, which will provide much needed information and resources to millions of people around the world who can learn about topics as varied as diabetes and cancer, and which should become a pretty good business as well. Four months ago we launched our Spanish edition, targeting Spain, Latin America and the US Hispanic market, followed last month by our Portuguese edition, targeting Brazil, and we will be launching in more languages and markets in the very near future.