There is a famous story about Steve Jobs meeting with the founders of Dropbox, Drew Houston and Arash Ferdowsi, back in 2009. At least I think it’s a famous story - it was written up in Forbes after all! Apparently, when Mr. Jobs’ pitch to acquire the then fledgling company failed, he told them: “you are a feature, not a product”. What did he mean by that, and why should it matter?
The idea of what Mr. Jobs was trying to convey is that some companies build solutions that solve a specific problem (or several problems), while others build technologies that make other solutions better. To solve a business problem, you need to offer a whole product. The concept of a whole product has been around since the 1980s, and it means that it provides everything you need to solve a problem. Not every technology that gets sold is a whole product.Take DocuSign as an example. The flagship product handles e-signatures and DocuSign built a $2.8 billion empire with a $12 billion market cap. But nobody has an e-signatures problem. Electronic signing is but a small part of most agreements, whether those are job offers, mortgages, non-disclosure agreements, and myriad other types of agreements. Each agreement type represents a business problem, and e-signatures are a feature in the solution addressing the problem. A very useful feature, but a feature nevertheless!
Docusign (apparently that's the new way of writing it) was a typical feature company until it eventually expanded into contracts management by acquiring Seal Software in 2020. The recent announcement of the Intelligent Agreement Management platform and the just announced acquisition of the workflow platform Lexion are logical steps towards becoming a whole product.
Speaking of workflow, there has been a debate going on for two decades about the differences between workflow and business process automation (BPA, aka BPM, aka DPA) . The short answer is that workflow is a feature, always embedded within a business application, while BPA is a whole product, usually used to orchestrate processes across applications.
Machine learning too is a feature, delivering value while embedded into business applications to do things such as estimate travel time, recommend spare parts to bring to a repair job, detect fraudulent transactions, or alert drivers to put down their phone. Artificial intelligence, on the other hand, could become a product, especially once someone builds an AI like C-3PO, J.A.R.V.I.S., or HAL9000.
However, there are some software solutions that can be either a feature or a product. Examples include office suites (i.e., Microsoft Office 365, Google Workplace), collaboration tools (Salesforce Slack, Microsoft Teams, Zoom), and file sharing (Dropbox, Microsoft OneDrive, Google Drive). These solutions can be a product when they are used to solve the employee productivity problem, but they are a feature when they are used to solve a problem that requires integration with other applications (i.e., Zoom integrated with a Call Center application). The first thing that Salesforce has done with Slack after acquiring the company was to embed it into existing Salesforce applications – treating it as a feature.
Generative AI too can be either a product or a feature. It can be used as a productivity solution (i.e., for research, copywriting, or translations), or as part of another application (i.e., contract management, search engine marketing, or…gasp…a customer service chatbot). OpenAI recognizes that and packages the technology either as ChatGPT, which is the product sold to users and priced by user seat. But the OpenAI technology is also available as an embeddable engine with an API. That’s the same technology but packaged as a feature and priced for ISVs by counting usage.
Does it matter if software is a product or a feature?
It matters quite a bit for the go-to-market strategy (GTM). Selling a product that addresses specific business problems requires the value proposition for the target buyer in the target market. Such products are typically sold to business buyers, and successful GTM strategy must understand who they are and what problems they have. BTW, the business buyer for employee productivity is usually in the IT department.
Selling a feature requires a different GTM strategy. A feature is part of somebody else’s whole product and its addressable market. It requires playing by the rules of that market, following the solution vendor’s GTM strategy, augmenting its value proposition, going after its target buyers, and making sure that the feature delivers value no matter what moves the vendor makes. For example, if you build a e-forms solution for Sales Cloud, and Salesforce decides to shift its strategy from Manufacturing to Financial Services, you will be figuring out the forms use cases for Financial Services. The same is true for Zoom integration with Marketo and Dropbox integration with HubSpot.
Coming back to Dropbox after examining the feature vs product dilemma, is Dropbox a feature? Yes, Dropbox is a feature that is often used in integration with applications from Adobe, Autodesk, Canva, and Zoho. But like some of the other horizontal end-user applications, Dropbox can also act as a solution for the productivity business problem.
So, Mr. Jobs was only partially right. There, I said it.