The best of both worlds: combining innovation strategies from hardware and software development by Emily R. Batt

Emily R.Batt is a senior product manager at KAYAK with a background in mechanical engineering.On stage of PRODUCTIZED, Emily shared her experience moving from physical product development to digital product development. Her talk explained how the capital, scale and lifecycle in hardware and software drive the difference in economics between two worlds.

What does it take to build hardware and software products? How we can combine best practices from both to be better product people? To help you find the answers we’ve taken a look at some key takeaways from Emily’s PRODUCTIZED talk.


Two years ago Emily left the job she loved in mechanical engineering to learn how to build products for the connected world.As Emily moved from building physical objects to the software products, she discovered that even though the types of products she was building was very different, the approach to problem solving was pretty much the same.

At Kayak, she focuses on running different experimentations that involve a lot of data analyses. Building a product that is accessed by millions of people every day closely embraces working on machine learning, improving the algorithms, understanding challenges that the users might have and developing the commercial strategy for the company.

Problem solving and engineering is never about building the best solution, it is about building the best solution within a new set of constraints. Every solution is a tradeoff between impact, risk, and cost.

If you are looking for an ideal solution, that is something that could be very impactful. An ideal solution would have a very low risk of failure and very low cost. Unfortunately, these are the unicorns solution spaces that don’t really exist. What you see more often is this incremental solution, something that is not really impactful, very low risk, very low cost. It is easy to implement, but doesn’t really move the needle.


As a Product Manager, you should be always on the hunt for powerful solutions. These are the ones that are delivering the desired impact. Unfortunately, they all come with a pretty high cost and a pretty high risk factor.

Very often, if you want to “productize” your idea and go from start to finish, be prepared to incur a lot of costs. The goal of a good PM is to make powerful solutions look more like ideal solutions. The best way for doing that is iteration. Let’s build something small. Let’s test it out. Let’s get a feedback. If you try a lot of ideas with SVP and MVP, then you do eliminate the costs of Productizing the wrong product.

Iteration is a very important tool and looks extremely different for hardware and software products. The most differences are driven by 3 main factors: capital, scale and lifecycle. These are the areas that really distinguish hardware and software development.


Prototyping can deliver the most value in helping enterprises and companies get ahead of technical and product risk when developing a new product. SlideShare

If you ask a coder what do they need to start a product, the first thing you will hear is “a laptop”. And if you ask the same question mechanical engineers, you will not get even two matching answers.

Whereas software can reach a wide base of users with very low capital overhead, hardware requires investment in material goods to produce and distribute.

So, how do you reconcile those differences? How can you play in a hardware space if you need different type tools upfront? Where we can see the gap closing is with rapid prototyping. Fortunately, makerspaces, fab labs, and crowd-funding are lowering the barrier-to-entry in hardware, making the capital requirements more analogous to those of software. The growing number of hardware startups in recent years reflects this sea change. The talk included a study of a startup called PharmChk, which utilized low-cost, rapidly-prototyped parts to build a device capable of evaluating, with high accuracy, pharmaceutical quality in developing nations.

Prototyping can deliver the most value in helping enterprises and companies get ahead of technical and product risk when developing a new product.

Where do these two worlds can combine? 1) Software can really learn the value of the hard tool from hardware 2) Hardware stone needs to learn the importance of lowering the barrier to entry.


KAYAK flight searches from Florida IP addresses before Hurricane Irma. SlideShare

Following product launch, product growth and scale influence profit curves. Rapid user acquisition is lower cost in software, as shown by an example in which an HBO app built from a single laptop sustained a million users by passing the processing demands to their devices. Whereas software products are expected to scale dynamically with user demand, hardware products are constrained by the number of units produced. For hardware it is difficult anticipate the demand for their product. In reality, hardware faces the threat of commoditization that software really doesn’t.

On the other hand, in hardware you can easily anticipate the maximum number of users based on the amount of units you’ve produced. In software you need to act more elastically. An example from KAYAK shows that good software products can respond to these changing demands.

What the two worlds can learn with regards to scale is that hardware needs to be a little bit more modular with how they scale. Software can learn the importance of improving feature quality for feature quantity


During the development lifecycle, the role of iteration varies for physical and digital products. Iteration is a powerful tool for mitigating technical and product risk. Both start from the same place: understanding the problem validating the customer, and brainstorming. But from there, the development cycles look very different.

The build/measure/learn cycle is well-established in software, and rapid prototyping similarly lets physical products reap the benefits of early user feedback. Nevertheless, once a physical product goes to market, updates to form or functionality can be costly — MVPs are therefore less practical in hardware than software, and most physical product iteration happens early in the lifecycle. Software, on the other hand, can continue to evolve continuously in real time as data reveals patterns about user behavior.


In the product development process if you can move the learning early up in the process, bring iteration into your product development cycle, you can get ahead of whatever your limited resource is.

Value is really unlocked when you can keep the cost of trying unusual ideas low, because when you are trying crazy ideas it is more likely that you are going to land on those powerful solutions.

A LiveSketch of Emily’s talk by Live Sketching

Access Emily’s slides presentation below:

About Emily R. Batt

Emily is a Senior Product Manager at KAYAK, where she develops products that drive significant revenue for the company and deliver traffic to major travel brands.Trained as a physicist, she has extensive her prior research experience is in big data at the Dana-Farber Cancer Institute and the MIT Media Lab. Previously, she was a mechanical engineer at Boston-area design company Fikst, where she built core technology for medical and consumer companies.

About Productized Masterclasses

The Productized Masterclasses are 2 days of hands-on masterclasses and insightful keynote speakers. On 27 & 28 May you’ll enjoy 4 masterclasses of your choice, get practical tips, and network with your peers. Come and meet Dan Olsen, Kandis O’Brien, Radhika Dutt, Ken Sandy, or Daniel Zacarias, among many others and get ready to be inspired to learn more about Enterprise Product and Consumer Product! SAVE THE DATE — MAY 27–28 2021

Want to be a part of the Productized community? Sign up for our newsletter here.

Through workshops, courses and coaching, we seek to contribute to the relationship between people and their organizations to build better products. Welcome!