This is the fifth post in my series on the future of work and labor. Here are the earlier posts:
- Macro Trends
- Increasing Productivity Opens New Opportunities
- Commoditization of Labor
- New Forms of Organizing
In this post, I will focus on how companies can make better decisions that harness the macro trends.
New technologies and lower costs allow companies to move quickly and outsource more tasks. This enables rapid growth but also helps competitors. The decisions around outsource vs in-house and standard vs custom are important and complicated; pitfalls are everywhere. Teams need to be extremely thoughtful if they want to gain and maintain a competitive edge.
First, some clarity around the terms I use in this post:
- Outsourced: work done by contractors or freelancers, not by employees
- In-house: work done by full-time employees, inside the firm
- Standard: generic, commodity components or processes; off-the-shelf rather than bespoke
- Custom: made specifically for the firm, either from scratch or by modifying an existing solution
Outsourced vs In-house
Outsourcing has been a major theme in my posts. As reputation signals improve and transaction costs decrease, outsourcing is easier and more productive. Freelance platforms speed up the search process for finding a contractor. Better collaboration tools make freelancers more productive in less time. For many tasks, the cost of using freelancers will decrease as just-in-time education improves and more individuals shift from full-time to freelance gigs.
Organizations contemplating outsourcing have many factors to consider. At a high level, they can compare the total cost of outsourcing vs the cost of creating in-house. To start, the costs should include money spent on fees or salary, respectively. Beyond the basic expenses, teams should factor in transaction costs: even outsource projects might require employees to spend many hours searching for and managing the contractors. Firms should also consider probability of success - the expected value of an outsource project might be less than the same project developed in-house.
Outsourcing ranges from individual freelancers to large contractors. Unless they have an established personal brand, independent freelancers can be hard to find and assess quality. When teams finds great freelancers, they can use them over multiple gig and keep search costs low. Former employees can be great contract hires as they are a known quantity and are familiar with the team, processes, and products. Large contractors have a vast range in quality and cost. Depending on budget, quality, and time constraints, there are myriad options, loosely adhering to the classic adage of “Fast, good, or cheap - pick two.”
Other factors are relevant on a per-project basis. Here are some important dimensions:
How integrated is the project? If a project is entwined with other parts of the business then it is more difficult to outsource and transaction costs are higher. If the project is on the periphery or only loosely coupled then it is easier to outsource.
Do the skills exist in-house? If the project requires skills that the company lacks, it might make sense to outsource. However, if the skill will be in regular demand then a company might want to hire someone with the expertise or train an employee.
Is there ongoing maintenance? If the project needs to be continually modified or updated, or if it is likely to have many bugs, it might be worth doing in-house. It is expensive to pay contractors for every minor tweak and employees might not be capable of quickly modifying projects built outside the firm.
A related set of questions are tied to employee utilization. If salaried employees aren’t working at full capacity, the project may be done in-house with little additional expense. If employees are already at full utilization, opportunity cost is a salient factor. As a company considers hiring a new employee for the project, they need to consider the duration for which the hire will be fully utilized; a new hire that is only needed for a short period might be an unproductive expense in the long run.
Standard vs Custom
As an organization grows, it will use standardized components and products for certain parts of its business and develop customized solutions for other parts. Deciding when to customize is difficult - bespoke solutions can be fun to build and provide key differentiation, but they are also more expensive than commodities. Standardized solutions are quick and cheap at the beginning but may require tremendous customization down the road. The correct choice depends a lot on the stage of the product, the size of the market, and how central the project is to the business.
For a small team with constrained resources, the answer is easy: use standard components for everything except the core competency / competitive differentiation.
Before Product / Market Fit
Until a team knows that it has the right product, all that matters is learning and iterating. Unless customizing a component directly speeds up learning or iterating, a team should use standardized parts. For most web startups, this means using open source libraries and modules; building a custom server or framework is rarely relevant to answering the key question: have we built the right thing?
I’ve seen early-stage e-commerce startups hire a technical team or spend serious money on outsourced development on a website. These startups rarely compete based on their technology, don’t need anything tailored, thus are much better off using a Wordpress or Squarespace template.
Once a team has zeroed in on the right offering, they must fight the temptation to customize everything. If they have cracked a new market, they will soon face competition. It is imperative to focus on their competitive advantage. If there are strong lock-in and network effects, then user acquisition might take priority as competitors race to win users in a winner take all scenario.
At this early stage of growth, resources are still constrained and opportunity cost is abundant. A team should only customize if there is a clear argument as to why it is the best way to gain differentiation or accelerate the customer acquisition engine. The team should be clear about which dimensions matter for differentiation and advantage, and only focus on those.
In a classic breakdown of strategic advantages, the options are: operational excellence, customer intimacy (deep understanding of the customer), or product leadership. If a product is positioned as low cost (requiring operational excellence) then the team might be able to outsource customer service (once they reach diminished learning from customer feedback) but should keep logistics or manufacturing in house. If a company is differentiating based on high quality experience, then customer feedback should remain in-house but logistics and operations might be outsourced (so long as this doesn’t risk a negative experience).
The bottom line: if it’s core to the competitive advantage, keep it in-house; outsource everything else.
Once companies mature and have a sustainable advantage, they can look at doing more in-house; however, they still need to be wary of customization. With limited funds, resources spent on ancillary customization are resources that could have been used to extend a competitive advantage. In a fast moving market, advantages can erode quickly, so companies must be vigilant. Only companies with abundant resources and a large customer base can afford to customize at the periphery.
For example, Facebook has their own datacenters and has customized their server stack despite that tech being irrelevant to Facebook’s competitive advantage. Facebook doesn’t compete on operational excellence, it’s focused on product leadership and customer intimacy. However, Facebook has so many customers and so much cash that it can justify spending money and talent on non-core activities; a small marginal improvement across one billion users is worth the investment - especially when those funds wouldn’t be allocated to core projects - those are already well funded.
Don’t Customize Prematurely
Even if customization makes sense in the short-term, it still has its downsides. A customized process or codebase will need more maintenance than the commodity version. It will also take new employees longer to become productive as they learn the new system. With an open source codebase, a company will benefit from improvements made to the main project and can hire employees who have prior experience with it.
If a market is still in flux or a company is still learning, the company will continue to make significant changes to the products and processes. It may be much easier and painless to swap out commoditized components then to continually customize an in-house effort or completely scrap bespoke developments.
While speed and learning are key factors in deciding in-house vs outsource and custom vs commodity, an important aim is to enhance option value. Teams should make decisions that leave them the most future options and flexibility. This will enable them to quickly apply new learnings, reduce redundant work, and change direction with ease.
Fluid Teams, Porous Boundaries
Organizations, regardless of size, should aim for fluid teams that collaborators can join and leave ad hoc. This is so teams can quickly scale up to tackle new problems and talent can reallocate as needed. It should be as easy for a contractor to participate as a full time employee - the boundaries between inside and outside the firm should become porous.
Toward this end, teams can make sure to explicitly document their assumptions, methods, and decisions; there should be clear guides to help new team members get up to speed. In larger firms, there should be simple processes for reallocating talent and for hiring and paying contractors. Some software companies might be able to harness volunteer labor by open sourcing their projects and soliciting contributors; however, these firms should be wary that managing an open source project is time consuming.
Competing in a Fluid Market
As technology improves and costs decrease - many markets will see a surge in competition. Lean startups will enter with new business models and either underprice incumbents or win niche customer segments that are underserved; this style of market disruption is thoroughly discussed under the umbrella of Disruptive Innovation. For startups - this presents myriad opportunities, for incumbents it is a constant threat.
For both incumbents and startups who will become entrenched, there is no easy trick to staying competitive. A full analysis is beyond the scope of this post but I’ll mention a few tactics. Staying focused on improving the competitive advantage helps but can still leave a firm vulnerable to entrants with radically different cost structures. Some firms can keep internal innovation high by spinning out small teams, tasking them with innovation, and empowering them to take large gambles. Companies can treat any new entrant as a credible threat and engage before much market share is lost.
In a fluid or volatile market, customer intimacy is incredibly important. A company that deeply understands their customers can better iterate on products that tightly fit customer needs; they’ll also be able to better market the benefits of those products.
If new entrants are constantly entering the market, brand and reputation become a critical source of sustainable advantage. If a firm earns a customer’s trust, that customer will be much less likely to switch to a competitor. If possible, a firm should aim for other forms of lock-in or switching costs. This is highly context dependent but might include long contracts, strong network effects, or proprietary data formats (yuck).
Many of the criteria that lead to outsourcing also leads to standardization; likewise aspects that make customization attractive are reasons to do work in-house. While there are strong correlation, there are also exceptions. Much of this post is summarized in this 2x2:
Tumblr is having issues with Image uploading - Here is the text version
Outsourced and Standard: Work that isn’t tightly integrated and won’t need to be modified
Outsourced and Custom: Highly specialized work that will provide a competitive advantage but doesn’t require full-time work
In-house and Standard: Easily integrated work that doesn’t drive the competitive advantage
In-house and Custom: Work that is tightly connected to the competitive advantage and that will be frequently modified
The ease of outsourcing and the abundance of standardized modules allow companies to learn and scale faster than ever; it also makes markets more competitive. Teams that choose wisely will have a huge advantage over rivals and greatly increase their odds of success.