Why consulting companies lose clients and how to keep yours

The global consulting industry is said to have lost around $30 billion of value in 2020.

It's more important than ever that consulting companies working on data projects retain and impress their clients.

Often, companies and personnel have their own style and measure for success. But this can make the partnerships between client and consultant run sour.

While you might think what you're doing is correct, it's essential to avoid a mismatch between what you perceive versus what the client sees. Otherwise, you might risk losing them.

Let's look at why this might be happening in projects that have a significant data element; how to mitigate adverse effects; and the tools and technology you can use to make things run smoothly.

Getting to the root of the problem

It's important to recognize that you shouldn't ignore the problem. Equally, you shouldn't commit to stop making mistakes. After all, mistakes will always happen in business.

Communications may vary depending on the role of the project, the stage of the project you're in and with whom you're communicating. And, given that many of our projects these days have a virtual component, communications are becoming even more complex and challenging.

But different, or a lack of agreed-upon, communication style may be hiding deeper problems than it seems.

So, with that in mind, let's dig into seven examples of where your communications might be going wrong and the best practices and technology you can use to solve the problems:

1. An incomplete implementation

If you do not cover all possible cases and inputs of a project, you risk handing over work that isn't up to scratch.

You can ensure an effective and efficient project delivery if you can:

Be a 'good developer' with automated testing scenarios.
Break the solution down into isolated components where one error won't corrupt other parts of the solution.
Translate technical issues into business terms to communicate where problems are and how they occurred.
Test on as large data sets as possible, which will be feasible with the right test data generation tools.

You can also free up expensive technical resources and focus consultant time on high-value work with a data management platform that will enable you to build reusable frameworks that hide the complexity and make your data consultancy projects more efficient.

2. The client never has an exact specification

Without well-defined deliverables and a timeline upfront, your client might end up disappointed with the results. If you're not clear about your plan, they'll start to fill the gaps themselves and assume the missing pieces of the specification.

You can mitigate this situation by creating a transparent process that:

Delivers fast iterations that shows data as soon as it's available.
Works on entire data sets, not just samples which might not contain all the unexpected outliers.
Documents interactions and new findings in a ticketing system.
Calls ad hoc clarification meetings when unsure.

Full automation of the entire process allows you to manage every step of your process in a single, transparent workflow. An automated workflow also means you can re-run the whole process as often as you need - an important part of being able to show results when you're iterating quickly. You can also establish key data metrics and integrate them effectively so that you can work toward a specific business goal.

3. Unexpected runtime errors

Unexpected runtime errors that produce incomplete or corrupt data can put a spanner in the works of a project. Unfortunately, large scale software is too complex to be bug-free, no matter how much testing you do.

You can better identify and localize errors by:

Designing for retries and graceful failure – some errors will heal themselves over time and can't be anticipated. The key is to report problems and retry.
Architecting towards a highly distributed and reliable infrastructure to avoid cloud fragility from fragmentation.
Conducting stress tests that verify the stability and reliability of applications.

The aim is to identify errors, find out what's going wrong, and tackle them head-on. Not to try to 'catch them all' and continue as if nothing has happened.

You can ensure data quality and design for bad data with a robust data management platform. This will allow you to automate bad data identification and correction, as well as manage rule definitions and reports on data quality.

4. Uncaught regressions

One change in the project pipeline can affect multiple and unexpected parts of the projects. This can quickly escalate if the client doesn't know anything about it.

You can troubleshoot any regressions in the pipeline if you:

Design reconciliation reports into the solution.
Define metrics and test scenarios with the client.
Collaborate on impact analysis - how will this affect the client?
Keep regression test results from previous runs to be able to compare and communicate issues proactively.
Name assets and processes in business terms to quickly communicate what works and what doesn't.

Suppose you remove the risk of miscommunications and give your clients a common language everyone can share. In that case, you'll be able to far better balance the business and technical elements of the project. This means you can update the client on any changes in the pipeline that they should know about before it's too late.

5. The consultant is overly optimistic

In a new, exciting project, it might be tempting for consultants to overestimate their ability to deliver technical solutions (a harsher description might be 'lying to get the win').

This means the resulting inflated scope will impact performance expectations.

To avoid optimism when selling and silence when reality proves to be too tricky:

Be realistic and upfront about potential complications and delays.
Do your due diligence on problematic areas and communicate them to the client.

Optimism is only practical when paired with realism. Exaggerated project predictions without the realistic planning to support them is a one-way street to underperformance and an unhappy client.

One way to help ensure you can deliver on what you promise is to build a library of reusable templates and accelerators, so for similar projects you don't need to build the whole thing from scratch every time but instead can just focus on configuring the specifics of that particular project - effectively giving you a head start on new projects.

6. Too much or too little project management

Consulting is a people-centred profession, where your ability to help others will determine your effectiveness in establishing a long-lasting, positive relationship with clients.

But it's essential to get the balance right. Too much project management can cause delays (everyone's had experience of unnecessary hours-long meetings when an email or a call would do), and too little can leave too many questions unanswered.

Both of them might lose your clients.

You can address this situation with:

A proactive approach to communicating changes and problems.
An offline source of communication for clients to look at in their own time.
A regular update on the project to avoid constant check-ins.

Fast, regular and proactive delivery helps clients feel involved and informed. A data platform designed for rapid iterations allows you to quickly implement changes - and deliver interim results - so your clients don't feel left in the dark.

In addition to this, look at investing in applications such as Trello, Slack or Monday.com. These will ensure there's a record and a steady flow of information for the client to look back on, and contribute to, when it suits them.

Make your data consulting projects faster and easier with CloverDX

7. An incorrectly deployed solution doesn't perform as expected

Deployment doesn't always go as expected. If there's an issue, the functional code will either work slowly, produce incorrect results, or not work altogether.

To avoid this happening, make sure to:

Deliver infrastructure as code, e.g. Docker.
Write detailed specifications.
Have a pre-configured solution to avoid client struggles in setting up their infrastructure.
Retain control over deployment if possible.

Find yourself a platform where you can design, automate, operate and publish data at scale. One that fits your client's environment and allows you to orchestrate complex deployment solutions.

Communicating with clients in their language

Consulting can be challenging. Quite often, you will face situations and problems in which your ability to manage client expectations, identify a risk-triggering event, and maintain clear communication lines is vital.

But it can also be very rewarding if you make sure you're communicating with your clients in their language.

The examples above demonstrate the transparent, open communication style that will let your clients know exactly what's going on. With the right technology, you can plan, develop, test, document and communicate any errors and unexpected changes that arise.

The CloverDX platform can help give your clients visibility into how their project is progressing:

Deliver rapid iterations so everyone can see progress and the results of changes
Reduce miscommunications between consultants, development teams and business users with a visual interface, so everyone understands what's happening to the data
Automated error handling means it's easier to deal with bad data quickly, and easy-to-understand error messages mean the business teams (the ones who really understand the data) can be in charge of how to fix problems
Complete flexibility for developers with the ability to code whenever you need - so complex data or legacy formats can be handled easily

CloverDX can also help your consulting company build repeatable, reusable templates for common data jobs, so you can deliver projects more quickly, onboard more clients, and maximise your income.

Get in touch to find out if CloverDX could help you get to speedier delivery, less manual work and happier clients.

Case study: How a consulting company was able to deliver Workday migrations 3x faster