6.3 Lessons Learned

In the post-mortem, consultants synthesize what they learned about dealing with clients and with their fellow consultants throughout the project. The following lessons were (sometimes painfully) extracted from 35+ past projects.

6.3.1 About Clients

More context for this section’s content is provided in Video 3.1 – Lessons Learned: About Clients.93

Welcome to the Client Dance

Getting a project off the ground can be exhausting (starting from when the client shows an initial interest to the time when work can start in earnest). Consultants start at a disadvantage and the client sometimes uses this as a negotiation tactic, by keeping the consultant waiting.

Solution: consultants should be polite, but they should also respect themselves, their abilities, and their work. While it will always be true that consultants need clients (in order to get paid), the converse is also true: clients need consultants (in order to get the work done).

Beware the Scope Creep and Divergent Expectations

The client may start by asking for a little “something” which is not explicit in the agreement (a different font, colour scheme, etc.). These are not big demands, and the consultants may agree too do so (for various reasons).

Then the client might ask for a little something else (repeat the analysis with a slightly different dataset to reflect that new data has come in, say), and perhaps the consultant could run a few more analyses, and so on. On their own, none of these demands are “big deal”, but when all the demands are added up, a whole new project has sprung up from these little bits, without the client having had to pay for it.

Solution: consultants should leave room in the agreement for modifications, with the caveat that the workplan may need to be revisited (as would timelines and costs).

It is entirely natural for the client to want something other than what they originally agreed to – as consultants start their quantitative work, they may expose conceptual and knowledge gaps, which could then lead the analysis into unexpected areas.

The agreement must be adhered to; modifications are possible, but at a cost to the client.

What Clients Want vs. What They Need

For the most part, clients do not always know what they want, from a quantitative perspective, and so what they want and what they need is not usually the same.

This often comes about because a previous consultant sold the client on an approach or buzzword, or because the client’s competitors are doing something specific and they feel they should follow suit.

Solution: it is the consultant’s responsibility to offer advice on what the client needs, not necessarily what they want. This advice should be documented because the client might decide to disregard the consultant’s advice and go with what they want (over what they need).

If, ultimately, the client comes to realize that what they wanted was not what they needed, the documentation should prevent the consultant turning into a scapegoat.

Talk is Cheap

Some clients are very gregarious: they are full of promises and full of ideas when it comes to a project, they will engage consultants in the process, … but for whatever reason they do not respond to the proposal, or they won’t agree to a meeting, or they won’t make the data available, etc.

Solution: consultants should not start work (in earnest) on a project until an official agreement has been reached.94

Disappearing Clients

Some clients pull out of the process at some stage (after the initial contact, after the proposal sent, after the project has started, at invoicing, etc.).

Solution: consultants should withhold deliverables until contact has been reestablished with the client.95

Consultants should learn how to sniff out disappearing or flaky clients; one way to reduce the risk is by maintaining healthy and regular communication. Consultants may have legal recourse if a client disappears at invoicing time, but they should be ready for a fight.

Helicopter Clients

The opposite situation can also occur: some clients are micro-managers and want to be involved with every aspect of the project.

Solution: consultants need to learn how to sniff out helicopter clients early. The team lead should consider giving the client nominal work to do and get them out the consultants’ way.96

The team lead is responsible for sheltering the consulting team from this annoyance and may have to shoulder the brunt of the interactions with the client, and may have to appeal to the agreement if it contains clauses that clearly delineate the responsibilities and roles of each party.

Desperate Clients

In another common situation, clients sometimes turn to consultants as a last-ditch effort to save a project (or to shift the blame to an outsider). A desperate client is often identifiable by demands for unreasonable deadlines.

This toxic situation can quickly become unbearably stressful and taxing for the consultant team.

Solution: consultants should be clear about the scope, the objectives, and the deliverables, as often as required, and be ready to return to the proposal often to remind the client what has been agreed to.While consultants need to show flexibility to keep clients happy, some limits should not be crossed.

Dishonest Clients

While the previous three lessons could be chalked up to clueless (and possibly non-malicious) clients, the next one cannot. There is no sugarcoating it: some clients will knowingly try to take advantage of the consultants.97

Solution: at times, paying work might be hard to come by, but consultants still need to do their homework and see what there is to be found about the client from external sources before an agreement is reached – it is important to trust instincts.

Sometimes, the problems only appear after the contract has been agreed to. In that case, the priorities should be for consultants to protect their team (including themselves) by document conversations and collect a (e-)paper trail.

Consultants should avoid threatening to sue the client unless they are ready to follow through with the suit; contacting a lawyer as soon as a problem arise is a must.

Procurement Issues

The proposed project authority on the client side is not always the person who holds the purse’s strings, nor do they necessarily have the final say on procurement matters – they may be very interested in getting the project going, but company or departmental policies could “get in the way” and complicate the process.

Most client organizations have their own internal process to hire consultants; for (small-ish) private sector clients, the issues are likely to be minimal, but for larger private sector clients and public sector clients, there are rules in place to stamp out corruption and nepotism.

Procurement vehicles include: sole-source contracts, standing offers, expert advisory agreement, professional services supplying, etc.98

In many cases with contract value thresholds (sole-source contracts, say), clients sometimes try to squeeze a large project in under a small budget, because the only reasonable alternative would constitute contract splitting, which is disallowed (in the public service, at least, you cannot give two contracts for the same project).

Solution: consultants should avoid selling themselves short, not only because it will mark them as “easy marks”, but also for a more pragmatic reason: if the client can only offer $25K for a project and the consultants agree to do $50K’s worth of work, the result is a $25K shortfall, which needs to be covered from somewhere else.

Speaking Truth to Power

Organization have the tendency to trust outsiders over internal experts,99 so there could be instances when the client already has a pretty good idea of what the data is saying but they need a person who is external to the organization to relate it to stakeholders.

And it could also be that the client knows that whatever report will be delivered will be poorly received by the higher-ups and they do not want to suffer their wrath, so the consultant is brought in as a the bearer of bad news.

Strictly speaking, there is nothing wrong with this, but some consultants might not enjoy being set-up to fail (as they might see it).

Solution: it is better to have all the cards on the table so that both parties know for what purpose the consulting team has been hired.

Consulting Witches and Wizards

Quantitative methods are seen as mysterious for a large swath of the population; consequently, experts in the field are sometimes viewed as witches and wizards.

As practitioners of the “magic arts”, consultants and data scientists are often saddled with expectations that can sadly not be met. Quantitative methods (coupled with sound data) can achieve many remarkable feats, but consultants do themselves (and their colleagues) a disservice by not managing expectations of their abilities early (and often).

Solution: consultants should be clear and direct about what they (and their methods) can do for the client, and nip in the bud any delusions the client may harbour about the project’s outcome.

This is also in the client’s best interest, and will stop them from making grandiose promises to their stakeholders – promises that simply cannot be kept.

It is always preferable to under-promise and over-deliver than vice-versa.

Calendars and Deadlines

Time managment is difficult in general, but in consulting projects, it is also difficult because clients are not always forthcoming about their own internal deadlines and calendars. They may attempt to move the project deadlines to synchronize with changes in their own deadlines.

The clients may also have deliverables for the project (getting back to the consultants in a timely manner, providing the data by a certain date, etc.).

Solution: when they hand in progress reports, consultants should remind clients of the timelines for each remaining task, as agreed to in the contract.100

The proposal should also reflect the effect of the client not meeting their project deadlines. Having a clause that reads “task 3 will be completed 2 weeks after the data has been delivered by the client”, say, rather than “task 3 will be completed by October 10” makes it clear that if the client delivers the data on October 8, the consultants will not have to scramble to complete task 3 by October 10 – if the task takes 2 weeks to complete, the consultants should get 2 weeks to complete it.

Data Availability and Quality

Invariably, the data is not as sound as the client thinks it is. And very often, due to internal politics the data will only be available way later than it was supposed to be, which can hamper the consultants’ ability to complete the project on schedule.

Solution: at no point should consultants accept the client’s word that the data is “good” and does not need to be cleaned/explored.101 The proposal (and methodology) need to reflect that data cleaning and data exploration are essential to the project’s success, and that consultants cannot guarantee the work unless they have access to the data.

Dealing with Adversity

Even when all the stars are aligned, and the consultants did a top-notch job and the clients provided domain expertise and quality data on time, it remains possible that the analysis results will not be to the client’s liking – data does not bend to anyone’s wishes.

As it is possible that the consultants made errors along the way, the client may be in the right to ask for a re-do.102 Where it becomes problematic is when they ask for a refund and/or put your credentials in doubt.

Solution: the proposal should reflect the nature of quantitative projects, i.e., the data/methods do not always support the client’s hopes. Consultants should not enter in a contractual agreement with clients who do not accept (and agree to) this fundamental fact.

6.3.2 About Consultants

More context for this section’s lessons is provided in Video 3.2 – Lessons Learned: About Consultants.

Importance of the Post-Mortem

It is impossible to learn any lesson if nobody knows what the lessons are. The importance of the post-mortem cannot be overstated.

Solution: post-mortems should always be conducted, even when the project was a success. It might also be a good idea to do project components post-mortem: consultants do not need to wait until the end of the project to identify what went well and what did not for a particular phase. Lessons are learned continously.

Boom or Bust

It is often the case that consultants (especially individuals and small teams) go through periods of months without a consulting project, followed by short periods where everyone want you to work with/for them, which wreaks havoc on work-life balance.

Solution: to survive “boom” periods, consultants need stellar project management. “Bust” periods can be used for business development, or for research and continuous learning.

Protecting Yourself Against Unreasonable Clients

As discussed in the previous section, unreasonable clients will happen at some point in every consultant’s career. While this might seem to be a lesson about clients, learning to deal with them is actually a lesson about consultants.

A project going belly-up is not the end of the world, although it can certainly seem that way in the middle of the blow-up.

Solution: consultants should get access to lawyer and a support system before the first sign of trouble – it might be too late to do so after the fact. Perhaps more importantly, consultants should be something other than consultants, at times – the benefit of physical exercise, hobbies, volunteering, personal time, and so on has been demonstrated time and time again. Don’t look down your nose at these simple solutions.

Teammates as Hurdles

Team work is not easy. Sometimes it will feel like teammates are hindrances in the pursuit of the project’s success – they just do not understand what needs to be done or they focus on things that are considered to be superfluous or out-of-bound by other consultants.

Solution: whether the other team members are missing the boat or not, they (and by extension, the client too) need to be treated with respect at all times – there will be times when having them as part of the team will be an incredibly welcome development.

With rare exceptions, consulting teams achieve more than individual consultants (although that may be easier to do when well-defined consulting roles are assigned and adhered to).

Academia vs. Business World

As quantitative consultants, technical skills are in high demand – they are experts and want to provide the best possible solution to their client. But their ways of thinking are often academic due to their training; they often favour the general over the specific.

In practice, theoretical solutions are not always actionable – they cannot always be turned into useful insights for the client.

Solution: consultants need to remember that it’s not about them: “best results” in the consulting context means “most useful for the client”, not necessarily for the consultant’s publishing record – being right is important, but it is secondary to being useful.103

Selling Yourself Short

Many quantitative people sell themselves short in order to try to not intimidate their contemporaries or potential clients – we would strongly suggest not doing that. Consultants have skills, and their time and work is worth something (quite often, a very high something).

Solution: know your worth and be confident.

Being an Arrogant Blowhard

But there is a slim line between confidence and cockiness – consultants being aware of their skills and technical proficiency should not translate as them being insufferable colleagues. Sherlock Holmes, Gregory House, and Sheldon Cooper might be intriguing TV characters, but consultants who try this approach in the real world will soon find themselves deserted by colleagues and clients alike, and, ultimately, without work.

Solution: consultants need to remember that consulting projects are not there for them to showcase how smart they are, but rather to give them a chance to help their clients achieve something they could not do on their own.

There are tons of qualified quantitative people in the consulting ecosystem, and more are joining the fray every year – standing out for being arrogant or a blowhard is NOT a viable marketing strategy.104

Keep Your Edge

Nothing ages faster than a quantitative consultant and their expertise;105 methods do not suddenly become invalid, but they can easily become dépassé.

Solution: consultants need to stay up-to-date and continue learning new methods and approaches on a regular basis, while maintaining their qualifications/certifications.