Oz Analytics

8 Weeks Into Our Beta Launch

It's been sometime since my last blog entry. Apologies, but I have a good reason: I launched my beta website doclens.com. It's now a couple of months on and time to communicate what the experience has been like.

First off, what is doclens?

Our mission is to make data driven narratives by discovering insights from reports and articles.We do this by:

1. Using the online research and opinion published by thought leaders such as consultancies, think tanks and academic experts. We estimate 50,000 – 100,000 items are published annually in our beta target market of analytics and big data people in the financial services industry. We automatically monitor this research as it is published.
2. Our DocLens platform automatically reads, understands and analyses this research and links it to profiles of the people and organisations involved.
3. We then curate the information and publish the best 1% to our website – making it possible for you to keep up with best practices, be the most informed person in your organisation, and make smarter decisions. All in just minutes of your time.

We turn information overload into personal strategic advantage:

Doclens is a lean startup. We have 4 people involved - with only one currently working full-time on the venture. We are also self funded. So resources are incredibly tight, but as the lean movement has taught us, it is amazing what can be achieved.

So here is a summary of the first 8 weeks of Six Degrees of Data as a 'trading entity'.

The launch has been a success. Being a data scientist, I measure success against a set of short-term targets we agreed on prior to launching. These are:

• 100+ subscribers by the end of 2014
• 5,000 article reads within 8 weeks of launch
• Average visit time > 2.5 minutes.

I'm pleased to say that all 3 targets were exceeded.

Whew ...

What else can I say about the experience so far? Well here are some of the highlights, in no particular order of importance or time sequence:

• For 3 of the 8 weeks we conducted A/B tests using 4 social media networks to advertise, using a total of 14 different messages. Twitter has been the outstanding performer for us.
• That said, we have also been actively tweeting and posting about the content being published on doclens each day. Judging by the number of retweets, favourites, mentions and clicks we are getting, this is maybe even more effective than paid advertising. I'm not yet sure how I can test this in practice (given our resource and budget constraints).
• Our customers are of course central to everything we do and some of the organizations they represent are:

• There is no particular geographic concentration in our hundreds of users - at least nothing unusual that I can detect. The USA, Western Europe and Australia feature heavily, as does Scandinavia, New Zealand and Singapore. Our users come from over 80 countries now.
• 95% are data and analytics professionals. Good news as this is are target market.

So what hasn't gone well or smoothly? A few things, that's for sure:

• Social media advertising. LinkedIn, Google and Facebook have not performed well so far. Advertising on Twitter was about 5 times more effective than these other channels. I still don't know why although my 600+ followers on Twitter may be a factor. I need to run some more A/B tests to resolve this.
• Customer engagement. People just aren't commenting on the content we are posting. Perversly, they seem to love the content and we get lots of visits. Our most popular articles have now been read thousands of times each. The silence is puzzling and I will probably survey our customers soon to find out what is going on.

Longer-term (6 month) targets were also set and these are:

• 1,000+ subscribers
• 20,000 article reads
• Average visit time > 5 minutes
• CAC < 40 cents
• Launch of our analyse-your-own-document service.

CAC = Customer Acquisition Cost. In our case, the cost of advertising on social media.

So these are the basic measures we are using. They are useful to have and us help focus on the things that matter.

It has been a slog, with many, many hours expended. And many more to come. But it is fun, we are learning every day and I wouldn't be anywhere else at the moment. I'll let you know if this is still how I feel in about 20 weeks time.

Brass Tacks

'Getting down to brass tacks' is an American expression that means 'get to the important facts'. This is what I want to do in this post. There are two reasons for this. Firstly I just realized that I have now posted to this blog over 200 times. Secondly, I am about to graduate from the Founder Institute (FI).

Both are milestones I want to celebrate by continuing on from my recent blog entry Helping Hands. In short I want to answer as directly (and as tactfully as I can) the question: How has the Founder Institute really helped me?

On reflection, these are the key insights I have gained:

• Traction is not one of the important things to worry about. In fact, it is the only thing to worry about! This is especially true if you are seeking external investors. I also learnt that traction means as a minimum completing (and measuring) a:
• 1st month offering a free trial to customers
• 2nd month converting trial users to paying ones
• 3rd month measuring how paid users continue to use your product.
• Customers must be engaged. Even if you don't have a product to sell, it is vital that you engage with potential customers to gauge their reactions and understand their feedback. If you can, make this something you do every week, every day. Call 10 people/customers before 10 am each day is something I must aspire to do if I am to succeed.
• People will inspire you. Many really want to help you and will go out of their way to do so. It is a real pleasure to have discovered  so many people that want to help me on my startup journey. Each week it is humbling to be reminded of this and it is an important factor behind keeping my spirits high.
• Disappointment. There will be many and the adage that it will take you 10 times longer and be 10 time harder to build a company than you think is true. You have to be tough and resilient to succeed. Even then the odds are stacked against you. I reckon that a founder of a high-tech startup has about a 1 in 100 chance of succeeding. I may be overly optimistic.
• Just Do It. Success is not only being smart and having a great idea. Yes, you've got to work harder than any corporate job, but more importantly, you have to set clear practical goals for yourself and your team. And these have to be extremely proactive and practical. Writing a report, or thinking about how to do something is no substitute for actually building a web page, or email/calling real people and achieving something concrete.

 None of the above is 100% of a surprise to me but what the FI has done is to drive home how vital each is. And maybe that is the most important lesson learnt so far: to rigorously focus on the customer, meet their needs. If you can do that - and 100 other things -  then you will make money.

Sounds simple, doesn't it?

One final comment. If you are considering starting your own high-tech startup, then I recommend the FI program as an option.

Before starting the program I did a lot of research on alternatives. I chose the FI because it was very practically focused on building a company, extremely cheap for what they seemed to offer and it was part time.

Now on the cusp of graduation I am happy (and relieved) to confirm that the FI has lived up to its promise. It has been a fantastic experience. One that I will remember for the rest of my life.

Come what may.

Helping Hands

nanos gigantum humeris insidentes

Regular readers already know the journey I am on with the founding of Six Degrees of Data and our work building a new way to read those documents that we should be reading but don't have the time for.

As a part of this journey I have been completing the Founder Institute's 2014 program in Sydney. The reason for this post is to explain a little about why I decided to follow this program and what the reality has been.

I'm still about 4 weeks away from completing the program, so I can still 'fail' and this should be thought of as an interim report. I will try to write again after the program ends. I may have a different message then.

I chose the Founder Institute (FI) because research showed that it was the startup accelerator that had the following features I wanted:

1. Internationally recognized and established
2. Intensely practical and hands-on, i.e. you actually build your company during the program (as opposed to talking about building ... )
3. The wide range (and number) of mentors involved in the program - and more are accessible through the FI network
4. It is a part-time program (so I could keep working)
5. Extremely cheap (for the potenial value delivered) both in terms of $$$ and equity sharing. It's a bit complicated but the bottom line is that the founder(s) give up 3.5% of their equity
6. Timing was good with a late May start date
7. It was local (i.e. in Sydney).

It may be that your needs are very different from mine, so the FI may not be as good a fit as it has been for me (so far). There are a lot of other programs out there so do your research - and I'm happy to talk to anyone considering their options. Drop me a line. 

The reality of actually doing the FI has, of course, been different - but not as different as I'd feared. On the whole the FI has lived up to expectations and this has been a very pleasant surprise for me. After all, teaching entrepreneurship is something a lot of people are very skeptical about. Put like that, I have to agree but I think this misses the point of the FI program.

I agree that you can't teach entrepreneurship as such. It has to come from inside of you - and you have to want it bad because chances are success is going to take a lot more time than you imagine. Just speak to any seasoned entrepreneur if you doubt me.

What you can do is accelerate the progress entrepreneurs are able to make. That's the true wonder of the FI and why it really is an accelerator. With less than a month till I complete the program, I already look back and can see that I am a lot further along than I would have been.

This is partly because of my own endeavours. I have worked hard and the 'part-time' part of the program often feels more like 'full-time'. Beyond this, what has been critically important for me has been then input, feedback and support of the mentors who attend every session. These are all seasoned and successful entrepreneurs who turn up each night and generously share their war stories, wisdom and contacts.

The best parts of the course may be the long sessions in the bar afterwards. The mentors give their time freely and I just hope that they know how appreciated they are by all us program founders.

Their participation is the backbone of the program. It sounds a bit clichéd as I write this, but I have heard more true words of wisdom and been given more business insights in the last few months than during all of the last 5 years I spent working as a senior executive in a global media company.

I can hardly believe it myself, so I won't hold it against anyone who doesn't believe me!

There is a big proviso to all of this and that is that success in the FI requires that you are either well prepared (in terms of developing your product) or able to move at the pace the program requires. Laggards will almost always flounder and have to drop out. To be fair to the FI, they warn you of this and the 'rules' are pretty clear up front.

The fact that only 5 companies remain of the 28 people that started the program, attest to this fact. Some dropped because they were simply unprepared for the pace of the program, while others could not satisfy the program mentors that their ideas would be enough to sustain a company.

Add to that the fact that you have to pass tests in order to get into the program in the first place. So the odds of completing the program are against you.

This is all to the FI's credit as the 'real' world is just as harsh, with 90%+ of tech startups failing in their first years. It may bruise your ego, but failing fast has got to be better (cheaper) than failing slowly. Right?

Plus there is the smugness of still being in the program.

That's a nice little boost to the ego at a time of great stress. I can hear Ben's voice in the back of my head: Whoo-hoo as he would say ...

What I Learnt Pivoting (so far)

Pivot:    (verb) turn on or as if on a pivot. In a startup it is a strategy for iteratively searching for a repeatable and scalable business model

At christmas we reviewed our strategy based on 6 weeks of feedback from our early adopters. The decision was to pivot and change Bridger from a mobile solution to a desktop one. Another 6 weeks have passed and I have learnt a thing or two that I feel it is worth sharing.

The good news is that there is good news. While pivoting is scary, we all feel that we have a better product now emerging. The consensus is that we are gaining:

1. A product that is more accessible for our users (i.e. it's on the right platform)
2. One that works with a single drag and drop movement
3. Has some very cool new features
4. Presents results more clearly.

We think that we will resume beta testing in about 6 weeks time. So the pivot is looking to have taken less than 3 month's to achieve (assuming no hidden disasters). The mobile version of Bridger took about 18 months to develop from the prototype stage. So the pivot will take only a sixth of that time. Is that fast enough to be considered lean? No idea, it is what it is. Certainly we are continuing to operate on the whiff of an oily rag so any progress measured in weeks is a plus.

At the start of this blog, I said that pivoting is scary because the lean books and websites 'pushing' pivots don't deal with the emotional investment you make in the original concept. It is your baby and as any parent will tell you, criticism of your child is hard to take and even harder to act on.

Although at 6DD we are now all hard at work building the new desktop Bridger, there is also a feeling of of inertia that is still to be fully overcome. Maybe inertia is the wrong word. It's more like regaining momentum after stopping the original strategy and switching to the new.

That's what I've learnt so far and I don't think we are unique in this. So if you are considering a pivot then I recommend that you spend a significant effort minimizing the emotional impact on yourself and your team. It is an intangible element that can suck the life out of your venture if you don't manage it well.

After all, if the empirical evidence tells you to change tack, what other options do you have?

One Month On

It has been a hectic month since we launched Bridger. Hectic and exciting with some real highs and lows. This post is to give you a little insight into the life of a startup over our first month as a company with a real product.

So what's happened and perhaps most importantly what's been learnt?

The first bit of good news is that our users have found a number of our features very cool, and some unique. The consensus list of Bridger 'good stuff' (in no particular order) is:

• User selects their own documents
• Supports multiple document types - and PDF support is unique to Bridger
• Understands a wide range of subjects.
• An extensive number of semantic targets are identified giving a rich picture of the document contents
• The mark-up of detected targets in document summaries and the full text is very useful
• Our dynamically generated profiles are unique
• It's mobile 
• It's multi-platform
• It's free - costs nothing and there are no ads

So we are all over the moon with this initial feedback from our users.

Unfortunately (and also inevitably) this is not the end of the story. There are also some issues. These include:

• Users are limited to selecting documents from Dropbox or webpage urls
• There are too many steps to get documents analysed
• Only short documents can be processed - some people want to process whole books
• It takes too long to get results - 10 seconds seems to be the limit
• Results have errors that make you doubt all results
• Smartphone screens are too small
• Some people don't understand what Bridger does
• A lot of users don't know that profiles are available
• The sixdegreesofdata URL is too long, and requiring user ids and passwords make accessing Bridger ATH (all too hard).

A few things to take on the chin but OK, the team can take it but the lack of instant fame and fortune is disappointing ...

Now some of these things are inherent in the nature of semantic analysis and linked open data. Others are 'fixable' through either the user experience (UX) or how we explain (market) Bridger.

In a way, we planned for this and the lean principles have prepared us for an immediate pivot of Bridger to address a number of these issues.

So stay tuned as we work on a radical redesign to the UX.