Software is quickly becoming integral part of human life as we see more and more automation and technical advancements. Just like we expect car to work all the time and can't afford to break or reboot unexpectedly, software industry needs to continue to learn better way to build software if it were to become integral part of human life. In this course, you will get an overview of how software teams work? What processes they use? What are some of the industry standard methodologies? What are pros and cons of each? You will learn enough to have meaningful conversation around software development processes. After completing this course, a learner will be able to 1) Apply core software engineering practices at conceptual level for a given problem. 2) Compare and contrast traditional, agile, and lean development methodologies at high level. These include Waterfall, Rational Unified Process, V model, Incremental, Spiral models and overview of agile mindset 3) Propose a methodology best suited for a given situation
Lean Startup
Loading...
来自 University of Minnesota 的课程
Software Development Processes and Methodologies
134 个评分
从本节课中
Software Development Models: Agile and Lean
In this module, we will learn about agile and modern techniques of software development. We will also learn about techniques that helps you learn faster about your market, your users, and your product.
与讲师见面
Praveen MittalAdjunct Professor
College of Science and Engineering
Kevin WendtTeaching Specialist
Department of Computer Science and Engineering
In this video, we are going to learn about a concept called lean
startup and it provides a very unique approach to software development.
And the basic idea behind this approach is,
how can you learn faster about your market or your user need?
So a lot of software development models out there,
they quite a bit mostly focus on the delivery of the software.
This model actually focuses a lot on understanding or fast learning about real user need.
And so let's see what it looks like.
So the lean startup model or lean startup concept was made popular by
Eric Ries with his book "The Lean Startup"
and primarily it was geared towards the startup companies.
But now, it has been used by a lot of industry,
at least the concept is very popular in the software industry.
So, as you can see,
the model is very straightforward.
It's a cycle of build, measure, learn.
And then again, keep doing this cycle.
And so that's the basic idea behind the software development.
But there are a couple concepts that makes it really powerful.
And the first one is how Eric Ries says about validated learning.
And so here is what Eric says,
"We must learn what customers really want,
not what they say they want or what we think they should want.
We must discover whether we are on a path
that will lead to growing a sustainable business.
And so, what he's talking about here is validated learning.
So you can learn about things by asking someone,
so you can just say, hey,
I'm planning to build this functionality, would you use it?
And people will say, yeah,
sure, I will use it.
But then what he's talking about is validated learning
where you actually collect data to find
out whether somebody is going to use your feature or functionality that you are building.
And so that's the concept number one is validated learning.
The another idea is to complete this cycle as quickly as possible.
It's all about time.
How quickly can you do the cycle so that you can iterate over it
and reduce the waste or reach your market in a faster way.
The third thing I want to talk about is a way to think about this lean startup concept.
So whenever you have an idea that you want to
build or something new that you want to launch,
a lot of times, there is assumptions about
that model that has to be true for that model to be successful.
And so, in lean startup concept,
you think about those assumptions and then you talk about what metric,
what can I measure to validate or invalidate my estimation.
And then once you have that metric,
you think about what experiment can I give,
can I do, and I made it red color intentionally to say they're just an experiment,
it's not even a product,
it may not be even a product.
So you do that experiment to get the metric and
that metric will validate
or invalidate your assumptions and then you can think about what are the next steps.
So, it has the concept of validated learning because you're doing
an experiment and getting a metric and then you are doing this as fast as possible,
and that experiment has to be such a way that you
don't spend years to learn about that metric.
So, let's take a couple examples to learn about this concept.
So you may have heard of zappos.com.
It's a website that sells shoes online.
And so when the founder of zappos.com,
part of this idea of selling shoes online,
he was thinking, will people actually buy shoes online?
And he wasn't sure and he didn't want to spend or invest
too much money creating an inventory of all the shoes and then start selling.
So what he did is this.
He went to a local store.
So he went to a local store with his camera and then took
pictures of all the shoes in the store and
then uploaded those pictures of those shoes on
a website and basically made it an online store for people to buy.
And then he just said,
let's see if people buy.
And so pleasantly surprised,
people bought shoes from the website and then it became very popular.
And then he started talking about inventory and all that stuff,
and as you know that zappos.com was acquired by Amazon.
And so it's a very successful venture,
but as you can see how he started.
So now let's apply the lean startup model on this.
So when the founder was thinking about building this online shoe store,
he has an assumption that people will buy shoes online.
And so he thought of what metric can I
measure that can validate or invalidate my assumption.
So he thought of if if I know how many number of shoes sold online,
that would be a good metric.
And so what is the cheapest, not cheapest,
but what is the fastest way I can learn
about this metric or how can I collect this metric?
And he said, why don't I just create a website with shoes spec and see if people buy.
So as you can see, he had an assumption,
he thought of a metric,
and then he did an experiment to validate his assumption.
Let's take another example.
You must have heard of Dropbox.
So again, the company who created Dropbox was a lot of technical people.
And so, again, they had an assumption or they were thinking,
will people need something like this where you want to share
your content or you want to synchronize your files into multiple devices?
So again, what they did,
what Dropbox did is that they actually didn't build the product.
They just build the video,
and the video basically shows how it will work or what will
be a user experience if they were to build the functionality.
And then, on the bottom of their video,
they had a link or a place for people to provide their email address
to be notified when this functionality is available.
So again, pleasantly surprised,
overnight, they got thousands of emails saying yup,
I'm definitely interested in this kind of functionality and we would like it.
And so, Dropbox actually executed on it and you know where it is right now,
very popular and a very well known brand in the industry.
So again, let's apply our lean startup model onto this. What are the assumptions?
So, in this case,
the assumption was people would want
the functionality to sync files on multiple platforms,
and the metric that they thought they'll
need is number of people who sign up to use the service,
and the experiment they did is to create a video that will show how it will work.
So again, they validated their assumption after they did
this very fast and easy on not very fast and quick,
the experiment, to collect the data to validate their assumption.
Let's take one more example.
So third example is about a company called Buffer,
and so they allow you to share and schedule
your content on social network
or social network site like Twitter or Facebook and things like that.
So, again, they wanted to know if people will be
interested in it so they didn't even create a video or a product,
they just created a dummy page.
They created a dummy page which says,
"Here is what our service will do.
Are you interested?"
And then the next page shows that we are working on it
and here is an email that you can provide and we'll get back to you.
Then once they've validated that,
then they did another experiment which was,
hey, will people pay for it?
So what they did is they created a page between these two.
And so, if the user says,
yup, I'm interested in it,
then they showed a page where they say select
the payment plan that you are willing to that you would like.
And then when they click on the one on the payment plan,
then they will say OK it is still in work in
progress and we'll get back to you if you provide your email address.
They did that to round of this validation of their assumption.
So again, let's apply the MLB model,
see if that works.
So the resolution in the round one was would people actually want to do this?
Would they warn to schedule posting content on social media?
The metric will be number of people who click on the link to show the intent,
and the experiment was to create a dummy page to show people for people to show interest.
Once they got this cycle done,
then there is another cycle of the same build,
measure and learn cycle,
which is would people pay for the functionality and their metric
was number of people who will click to the second page which is the price plan.
And so the experiment was to include another page which will
show the pricing plan and for people to actually select the pricing plan.
So again, they did a two round,
and so you can keep on iteratively making sure that you're going in
the right direction for your product in a very trendy fashion and quickly.
As soon as you see that something is not going to work out,
you can drop this idea and start with another idea.
So this lean startup style of concept really allows you to quickly learn
about your users or your market.
So, one question that comes to mind is that all of the examples that I
gave are for kind of a startup company.
Does it even apply to a big industry?
And so, luckily or whatever you want to call it,
I ran into this situation where I was paying for a service and in last five years,
the price I was paying or the cost I was paying for,
the service fee that I was paying was almost doubled.
And so I wanted to check and I want to talk to
the customer service for the company and say,
why did it go doubled?
And so, to find that,
I wanted to see invoice,
which was five years old,
and the invoice today and I wanted to get
those numbers so that I can show it to the customer service that, hey,
you guys have changed so much fees and it has almost doubled.
And so to do that,
I went to their website and I was looking at it and I saw a link.
It was showing me the current and the last year invoices,
but then there was a link to view older documents.
And when I clicked on that link,
it actually showed me this message saying,
"We are currently working on adding new features to
the documents page such as adding multiple years of documentation.
By clicking this link,
you're giving us valuable data to determine the demand for the document beyond two years.
Thank you for helping us complete this announcement.
A full-functioning page will be available soon.
So, clearly, whether they intentionally did it or not,
but this is clearly an idea or the concept of lean startup that the big company
used to make sure that they are working on the things that user actually need.
And so, again, let's apply the model.
So in this case, the assumption was there is a need to see all invoices.
The metric would be number of people click on the link to see the old invoices,
and the experiment was to create a dummy link to show the old invoices.
So as you can see, again,
it's a concept that could be applied to startup as well as in big industry,
big companies, you can apply this concept there.
So let's summarize.
What is the lean startup is all about?
So it's about validating learning and it's
about completing the cycle as quickly as possible.
Of course, it's an incremental and an iterative approach.
And then kind of believes in the short, very short cycle.
In terms of the predictive and adaptive cycle,
really obvious that it is very much on the adaptive side.
In fact it's on the negative side of the adaptive that it even goes beyond.
So in terms of pros and cons,
I would say it helps you learn faster and helps you build the right product.
And of course, for speed to market,
you can quickly go to the market by validating it.
In terms of cons, again,
may result in a rework and it requires you to experiment iteratively with your clients,
not just with some proxy
for the users but actual users and the client you experiment with them directly.
So in the examples that we saw,
we had just a link and then when somebody clicks on it,
it just says, oh, we are still working on it.
So there is a little bit of risk there that you're
experimenting with your users and your clients directly.
So where do you use this?
If you have a very doubtful business case or
a user need or if there is a lot of high probability risk,
then you can use this concept.
Of course, if you are building
a big product and you're just not sure about a specific feature,
then also you can use this concept to
validate it before you put too much effort into that feature.
So that's lean startup.
