Over the past few years, we’ve dealt with outsourcing in various capacities. I’ve been on the outside looking in. We’ve been a vendor, selling our services. And we’ve been a client, looking for competency outside rather than hire – or develop – such competencies from within. And there has always been, we’ve noticed, a trade-off.
Before delving deeper into this topic, it’s also important to stress that this question – to outsource or not – is a perennial question applicable to all sizes and sorts of companies. A startup bootstrapping its way to an MVP or a prototype may approach it with a certain decision matrix, while an MNC, a behemoth spanning multiple countries and businesses, might execute a different algorithm to arrive at the answer, but the question will always be the same:
Is it worth it?
The “it” here refers to the perceived sense of loss of control when you outsource a work. We say perceived, because in many cases, with proper guarantees and milestones, there is often little to no loss of control. Indeed, as many startups who’ve raised just enough funding to build up in-house technology teams will attest, life might just have been way simpler if they had not insisted on “control” over “guarantee.”
A hire is an act of intimacy for a company, after all. There are the search costs that cover advertising and shortlisting, and as often as not, there is a third-party vendor here as well, especially for the more senior roles. Then come the round of interviews, and that means productive time must be set aside for long discussions where the candidate and the company court each other, gauging each other’s fit, negotiating numbers and joining dates, setting down responsibilities, determining reporting structures and managing deviations from original expectations. That sounds complicated enough as it is, and it is for just one post!
Repeat as many times for as many candidates as needed until you have the team that you need.
Since the whole process is so time-consuming and, to use an expression we’ve heard many times in this context, invasive, there comes into it an element of escalation of commitment even when things don’t go as well as they should have. Such hires, whether at startups or later-stage firms, have to be trained on in-house processes and tools; they are exposed to confidential information, often with competitive advantages; interpersonal relationships are formed and re-formed with existing employees. The company’s DNA absorbs theirs, changing both in the process.
Which makes hard decisions even harder to make at the end of the day. There are companies we know of which dodder along with their technology teams simply because it is not in their culture to let them go or demand, with growing impatience, better results. There are companies who’d rather let their cash reserves run out on adding pointless fixes to their systems when what they actually need is a radical overhaul their in-house team is not qualified to provide. Not to mention, of course, the actual costs of separation: the lack of productivity during notice periods, the final settlements, the paperwork and the retrenchment of those who should not be let go.
Outsourcing, on the other hand, brings no such baggage with it. It’s perhaps – to use a rather crude analogy – as different to doing things in-house as dating is to marriage. When an outsourcing engagement fails, the costs are limited, controllable, the damage to the brand negligible; when an in-house team fails to deliver, costs can spiral out of control and the brand might take a serious hit (such as the signal sent out when a company summarily releases a large number of employees: Ah, cost cutting? Must be in trouble!)
Undeniably, there is a loss of control when you outsource. Or, to put it another way, the safeguards for performance are different with an outsourcing engagement. You must have done your due diligence and ensured that your vendors are reliable and operate within legitimate frameworks. You must be clear about your own requirements and not “wing it” as you go along. And, most importantly, never let the vendor feel that your project isn’t important or that you aren’t paying attention. A vigilant client is the best client for firms like us – it keeps us on track, and deliveries are less likely to be plagued by last-minute course corrections.
An ironclad service agreement is in both parties’ interests, guarding against delays, overruns, scope creeping, analysis paralysis and payment-against-deliverable disputes. A best practice is often to nominate a key, senior stakeholder to the project, someone who’s been involved right from the ideation stages to where the final roadmap has been established. Such a person would not need to go back to the others to revisit a question that has already been discussed; such a person is ideally placed to decide how quickly a question should be answered.
Most outsourcing engagements that fail, fail on these counts. Agreements signed in bonhomie might make for a happy moment at first, but leave a sour taste in the mouth at the first sign of trouble; on the other hand, a detailed agreement that leaves little to interpretation or chance might be a bit of a headache at first but saves you a migraine down the line.
Executing a project in-house, naturally, means there is no SLA, no contracts to be thrashed out, no points of law to be gone over by legal teams. It also means that there is less accountability, for a vendor gets paid only when he delivers whereas an employee has a fixed paycheck coming to him at the end of the day anyway.
In other words, if you are serious about the project and want to see it executed within your deadlines, find out who has the most pressing incentive to see it delivered.
Something else that’s often hampered in-house teams from delivering is the very fact that they are accessible to the rest of the company. Thus, a team hired with the specific objective of building and maintaining a complicated enterprise application might find themselves being requisitioned for minor tasks like updating the company website or running low-level fixes and building ad-hoc features. The organization might feel that it is getting its money’s worth, but at what true cost?
A vendor, on the other hand, is insulated from such pressures and demands. Additional work can be gauged objectively on the basis of effort, time, capability and remuneration; there is no blow-back on anyone if there is a refusal, nor will the existing bandwidth on the original project be affected because of a division of resources. The ideal vendor wouldn’t, shouldn’t, compromise the original work for bits and pieces that are offered if they are detrimental to the overall objectives of the engagement.
A study by Gartner a couple of years ago pointed out that the vendor’s limited role in decision-making as far as the product’s features are concerned is a major factor enabling faster development through outsourcing. Even when the vendor wears the additional hat of a consultant, as we often do at Pluto-men, there is a point beyond which we will not – and should not – belabor a point. At the end of the day, the company’s managers must still be the ones calling the shots.
When development is done in-house, there is still a possibility of vested interests and stubbornly-held opinions influencing the project’s execution. A miffed manager might refuse or delay access to information critical to development; a stakeholder might still try to tweak the requirements or sneak in features already discussed and rejected. With a vendor, the feature set is inked and fixed; an in-house engagement has no such structure, nor defence against such dark arts.
Even though we are vendors ourselves, there are still tasks – non-core competencies – that we outsource. For instance, for the videos we create, we retain the services of agencies or artists as and when needed for the voice-overs. We don’t stitch our company t-shirts – we order them custom-made by a vendor. It all boils down to this at the end: give it to the people who do it best, and just take care of the rest!