Kill Your Product – Why Sacrificing Your Cash Cow Can Be the Path to Growth
An article by Shane Evans, CEO of Zyte.
- Published:
In the tech industry these days, funerals for software are all too familiar and the graveyard of discontinued products is ever-growing. Whether it arises through company failure, M&A or market shifts, the decision to sunset software can evoke sadness, embarrassment, fear and resentment.
But killing your software can be a path to success. In fact, sunsetting your biggest product at its peak could be the move that unlocks a brighter future.
That’s what I did when I shocked my team by announcing we would deprecate the product accounting for 60% of our revenue. Here is what I learned, and why I think this bold move is sometimes necessary.
Piece-by-Piece Proliferation
All software is the story of laddering waves of new capability at the intersection of problem and opportunity.
My web scraping journey started in 2007 when I wrote Scrapy, a web scraping framework, to support extracting data from e-commerce websites. Within two years, my team used it to gather data from 4,000 websites.
However, additional challenges arose over time. When websites started blocking access, I wrote Smart Proxy Manager to route requests through a large list of IPs, manage them, and avoid getting blocked. Further capabilities were added as separate products, such as a residential proxy offering and the Smart Browser for large-scale browser rendering needs.
But the trouble with incrementalism is that one day, you wake up and realise your offering is really a smorgasbord of cumbersome point solutions.
Complexity Creeps Up
Servicing a suite of tools drains an increasing amount of time. As the task of modern web scraping grew in complexity, demanding several different approaches, we shipped products for each. But our stack became so complex that even our expert users lost time deciding on the optimal solution or responding to website changes.
Smart people, skilled at assembling pieces of a tech stack from disparate sources, won’t always complain to you about this sort of friction, because solving puzzles with competence is their job; technical challenges are business-as-usual.
Moreover, this proliferation of products was considered good practice at the time – every other vendor was rapidly adding more products, often with overlapping use cases.
However, when providing customers with a collection of isolated tools, many remain oblivious to the full range of options available. They often don’t realise when they have made a sub-optimal choice, and can fail to recognise possibilities beyond their immediate needs.
Rip It Up
The answer to our problem lay in combining our offerings in a single API that could address the whole web scraping stack, making optimal use of the infrastructure and avoiding the need for users to manage all that complexity.
But sometimes people become accustomed to the status quo. Product managers assumed the new API would be an add-on to our primary product, Smart Proxy Manager, because they could only perceive iteration through our existing product offering.
So, when I said, “Guys, we’re killing these products,” people were shocked. I announced that, in a couple of years, we wouldn’t be selling the standalone products anymore – instead, we would build a single, brand-new product, an all-in-one web scraping API, called Zyte API.
I don’t mind admitting, the team thought I’d gone crazy; people were unhappy. A year after the switch, however, we have seen a 15% increase in revenue from migrated users. Even though the new product is cheaper on average for the same workload, usage is up considerably as it can be used on a broader range of tasks.
Sunsetting Is Success
So, don’t mourn for deprecated software. Sunsetting a product can indicate a mature software category experiencing strong growth, momentum that has driven a creative explosion of diverse solutions, which now need to be rationalised.
A company’s willingness to kill a product shows that it is evolving fast enough to outpace its previous offerings, transforming standalone features into a larger, more ambitious vision.
You can expect to see a lot more software being sacrificed in the near future. AI is such a step change that it will prompt a fundamental rethink of many products, including in the web scraping field, where large language models will transform the ability to parse unstructured data.
A Funeral for Your Flagship
If you are coming around to the value of bidding goodbye to your main product, what are the main considerations?
1. Build Internal Buy-In
Your team may provide the most resistance. After all, staff are wedded to and care deeply about what’s on their plate right now.
Your job is to build confidence in your vision for the future. Build a coalition of internal support by showing how the medium and long-term benefits represent a bigger prize. I had to communicate the vision clearly, demonstrating how an integrated API solution would ultimately save us time, reduce costs, and improve the customer experience.
Unfortunately, you won’t always convince everybody, and you must still proceed despite some opposition.
2. Reallocate Resources Meaningfully
Stopping doing something frees up the resources to do something else. This is the fuel that gives your future room to grow. Embracing that opportunity means deciding to stop actively developing the outgoing product.
Had we not actively stopped new feature development on Smart Proxy Manager, staff would not have taken Zyte API seriously. This goes for sales as well as product teams – we had to stop selling our older product.
3. Take Users on the Journey
4. Build a Bridge to the Future
But it’s not just about communication. Technical customers’ anxiety about product deprecation is real and understandable because no one wants to be forced to write new code for something that works perfectly well. Offering backwards compatibility, as Zyte API did, can massively minimise disruption to users. A commitment to continuing to support critical enterprise customers will always go a long way to guaranteeing continuity.
Kill or Be Killed
Letting go of the past is the best way to embrace the future. Retiring a flagship product isn’t a sign of failure; it’s a commitment to innovation.
As we enter this new era of disruption, I wonder if companies will be willing to disrupt themselves before it’s too late.
