Chapter 8. Tailoring Your Work to Specific Departments

When you get good at communicating with data, opportunities will open up. You might well find yourself working on bigger and more diverse projects with new teams and departments.

The new people you work with will have new terminology, different data sets, and different stakeholders. As you’ve seen throughout this book, understanding your audience is crucial, and with new departments you may not know the full context of every situation. You can address this problem using the requirement gathering process you learned in Chapter 2.

When you step into new roles or new departments, you might even feel a bit of imposter syndrome, the sensation of self-doubt many people feel when they step out of their comfort zone and into new opportunities: Do I really belong here? Try to ignore this feeling: it will fade as you build more confidence over time by continually proving the value of your work. Just remember: if you are being offered new opportunities, you have earned them. If someone else believes in you, believe in yourself!

I have been fortunate to work with many different types of teams and departments in my career, and I have learned some of their needs and wants. This chapter is a bit of a cheat sheet for you. It won’t tell you everything you need to know—every organization is different—but will give you a starting point by noting some challenges that tend to arise when doing data work with departments like human resources, marketing, and IT, as well as dealing with senior-level executives. If you’re prepared for these challenges, you can start delivering benefits sooner and continue your good work.

The Executive Team

“Claire?” Preet pops his head into Claire’s office. “I have some news for you. The CEO liked that dashboard you made me. In fact, she’d like you to make her a dashboard so she can monitor the company’s performance. Are you ready?”

The first team beyond your normal department you will work with is probably the executive team, or ‘C-Suite’: the most senior level, which might include the Chief Executive Officer (CEO), Chief Financial Officer (CFO), and—in more and more organizations—the Chief Data Officer (CDO). If you are anything like Claire (or me), communicating to this group can be intimidating.

Claire seeks advice from Wang, a senior coworker who deals regularly with the CEO, Erin. “I’m nervous,” she confesses.

Wang nods. “I get it. It feels like you’ll be fired on the spot if you make a mistake, right?” Claire nods. “You must remember, though, that you are there because these people are looking for your insight and analysis to inform their decisions. Think of this as an opportunity to make Prep Air better.”

“Thanks, Wang, that’s reassuring,” Claire says. “Any advice? How do I get them to listen?”

Wang smiles. “For me, the challenges are time and scope. Think about what it’s like to be CEO. Erin has hundreds of opinions, emails and reports landing in her inbox every day, and she has to figure out how to consume all that information. Who should she believe? How much does each view sway her opinion? What should she try to solve or improve next? She has so much information coming at her that she can’t waste any time. So my first piece of advice is to communicate clearly and succinctly. Get your message across fast.”

Claire is already forming a plan. She decides to start with a quick overview that clearly indicates what the executive team should focus on. She decides to create a dashboard for Erin, starting with a landing page: the first thing the audience sees when they open a dashboard. It gives an overview of the subject, then steers the audience’s attention towards the most important areas. This way, Erin will be able to find crucial issues and information in the data quickly, instead of having to scan dozens of reports hoping to unearth them.

As Wang noted, Claire will need to put her effort into thinking about scope. What should the landing page include? How can she cover the whole company’s performance broadly, without missing anything important?

Claire doesn’t want to bother Erin with lots of requirement-gathering questions, so she decides to use the company’s key performance indicators (KPIs) as her guide. The executive team sets the company’s KPIs, which are important measures that, taken together, form a picture of how the company is performing in relation to its goals. Prep Air’s goals, as present in the last all-hands meeting, are:

  • Increase revenue

  • Increase profit

  • Improve customer experience

Executives spend a lot of time trying to understand the best ways to measure these key drivers; when they decide on the right measures, they designate those as KPIs. One way you might end up working with them is by having a lot of experience in one of these drivers. An increase or decrease in a KPI will trigger the executive team to ask what’s causing the change. They know that there’s a lot of additional detail underlying each of those measures. That’s where you come in: you are likely to have metrics and qualitative information that can help them understand such changes.

Let’s take a look at Claire’s landing page (Figure 8-1).

  Landing Page example
Figure 8-1. - Landing Page example

There are two areas that instantly jump off the page: Capital Expenditure and NPS. Note that Claire has used red for these—here, red doesn’t indicate a decrease or increase, but any change that is for the worse. The CEO can spot these and focus on them quickly, without having to dive into the details of other areas.

Of course, there’s a lot more there than just a number. A set of reports needs to sit behind each of the tiles on the landing page so the audience can click through to find supporting data and investigate the issue. If Erin clicks on the Net Promoter Score indicator, she’ll arrive at the screen in Figure 8-2, which dives into the details.

  Detailed page example for NPS
Figure 8-2. - Detailed page example for NPS

There’s a lot of information here, and pulling it all together involves a lot of collaboration across the business. The stakes are high, too: this information is going to provide the basis for decisions that could change the whole organization, so its accuracy needs to be spot-on.

Claire, realizing this, returns to Wang. “How can I make sure everything is right when I’m not familiar with how all of these departments work?”

“You can’t,” Wang replies. “You’ll need help.” Collaboration, he explains, makes the difference in these situations. “Here’s what you do. Draft some visualizations and share them with people in the relevant departments. They’ll spot your mistakes. They might also disagree with you about what the best measurements are or how to read the data. It’s all valuable, and getting those departmental perspectives is really the only way to know what you might be missing.”

Claire designs the landing page to provide the executive team with a useful overview of the organization’s performance. She submits her work and waits nervously. A few days later, she finds a thank-you email from the CEO herself. “I appreciate what you’ve built here. This makes it easy for me to identify areas of poor performance and find the information I need to take quick action. Great job.” Claire forwards the email to Wang, who congratulates her and adds, “Don’t be surprised if you find yourself doing this more often!”

Finance

After Claire’s success at providing the CEO with a dashboard, word spreads. It’s not long before she receives a similar request from the CFO, asking for a dashboard the finance department can use to track ticket revenue.

Finance teams require lots of timely and accurate data. They’re constantly analysing data sets on income, expenditures, and more. Claire, whose financial experience is limited to doing her taxes, finds herself intimidated once again. “The financial team has so much experience and expertise,” she tells Wang. “How can I help them?”

Wang smiles. “Let me tell you something. I’ve worked with many different financial teams, and there’s a grain of truth in the stereotype that all they want to work with is tables. No, it’s not everyone, but they like to see the data points as clearly as possible so they can dig in.”

“So… they hate charts?”

“I wouldn’t go that far, but finance people do tend to be skeptical about data visualisations. There’s a balance. They usually want to see the rawer data even if you also give them charts I’d lean toward tables if I were you—especially since they’re going to ask for reconciliation.”

Claire grimaces. “What’s reconciliation?”

“It’s when you compare the values you’ve created to values that you know are correct. There are different ways to do it, but it often comes back to checking values against what has previously been reported in tables. They’re going to use this for regulatory reporting, taxes, and statements to investors, so you’ll want to triple-check all of your numbers.”

Claire considers this as she ponders how best to share her message. If they want this much detail, how can she make sure the message comes through clearly? She’ll also need to make sure the audience can follow—and check—her logic.

Claire decides it’s important to convey the main message before adding detail—otherwise, it might get lost in a sea of numbers. She decides to utilize the ‘Z pattern’ (discussed in Chapter 6), placing contextual numbers and charts at the top of the page and then showing the detail further down. She creates Prep Air’s ticket revenue dashboard (Figure 8-3) with a table at the bottom of the view that allows users to validate the visuals and calculations.

  Financial dashboard with detailed table
Figure 8-3. - Financial dashboard with detailed table

The dashboard she creates is interactive: the financial experts can use the charts at the top to filter the detailed table below. This way, they won’t have to search through a large table to reconcile the values shown with known comparables. The charts themselves can act as a filter: they can simply click on or hover over the marks. For instance, if they click on “Paris” in the “Ticket Revenue by Destination” visualization in Figure 8-3, the table at the bottom of the dashboard updates with figures specific to that city (Figure 8-4).

  Updated table at from  fig_3__financial_dashboard_with_detailed_table when Paris selected
Figure 8-4. - Updated table at from Figure 8-3 when Paris selected

The charts that can steer your audience to the filtered tables can also offer more context. Charting makes the stories in your data stand out, as you’ve learned. The same is true when trying to reconcile data points.

Claire decides to ask for feedback, as she did with the executive dashboard, before finalizing her design, finance users are likely to have a detailed understanding of the subject and can identify potential outliers or mistakes.

Human Resources

Claire’s financial dashboard is a success. Before she knows it, Erin the CEO is back in her inbox. She wants all of Prep Air’s departments to have dashboards of their own. HR, operations, marketing, sales, IT—everybody wants one! Claire is thrilled to see that the whole company appreciates her data communication work, but she also knows she’ll need guidance on working with such a diverse range of departments. She schedules a meeting with Wang. “Can you give me some tips for each of these departments?” Wang congratulates her on her excellent work, and they dive in, starting with HR.

“In my opinion,” Wang begins, “the biggest challenge with human resources is the datasets themselves. You’ve got to be very careful with using and sharing sensitive personal information. Imagine how you’d feel if someone was careless with your private data! And of course you need to respect regulations like GDPR.” Data sets can contain many sensitive data points, he notes, and many of the most sensitive lie in the hands of the Human Resources team. Prep Air, like any organization, keeps records of every employee’s pay, age, home address, and number of dependents, to name just a few.

One common technique for visualizing sensitive data is to aggregate it, or show only summarized data. In other words, instead of showing individual data points, you might take information from five or ten individuals (at minimum), then use the median of that information as a data point. If you do this before beginning your analysis, it is known as pre-aggregation. Pre-aggregating your data makes it much harder to identify the individual people from whom the data is drawn. Although this grouping technique won’t give you exact accuracy, it will allow you to share messages more widely than you would be able to otherwise (Figure 8-5).

  Chart with aggregated data showing median salary per grade
Figure 8-5. - Chart with aggregated data showing median salary per grade

If your data is not pre-aggregated, take special care. When you are filtering data for multiple characteristics, it can be challenging to ensure that you don’t inadvertently leave individual people identifiable. In the grades shown in Figure 8-4, the triangle-circle icons show how many individuals were grouped together. The groupings for Managers and Team Members are sufficient that the median salary won’t reveal anyone’s individual details. You’d need to take care with the Executive’s details as there are so few of them.

However, if you broke each of these groups into individual departments, it would be easy to identify the salaries if you know who manages which department. In addition to potentially violating privacy laws, this could bring up morale issues, such as:

  • Individual employees could see their pay relative to that of their peers. If their peers are making more, they could feel undervalued and request more (or leave). If they don’t see salaries higher than theirs, they could infer that there is a limit to their growth in the organization and begin looking for opportunities in other companies.

  • If employees see disparities between different departments, this can create resentment.

  • If some individuals at lower job grades who have unique experience and skills are paid more highly than other employees at senior grades, this too could create resentment.

The chart does not provide any context that would help employees understand such cases.

One technique you can use to avoid this problem is to set a minimum count of individuals on the filters you use for each item shown, for example so that the value will only be shown if the group includes at least five people.

Wang advises Claire to account for how her audience might interact with her work to ensure that the messages shared don’t reveal individuals’ details or create interpersonal conflict.

Operations

“And then there’s operational data, which has exactly the opposite challenges,” Wang says. “Getting into the details is the only way to find out what’s happening, what can be improved, and what kinds of investments that will take.”

Of course, operations departments are all different, depending on what the organization does: operations might focus on servicing vehicles, teaching classes, or manufacturing products, for example. At Prep Air, the operational teams handle everything from cleaning planes to selling tickets to handling customer complaints.

Everything an organization does generates data. Communicating that data clearly allows managers to measure operational processes and identify problems (or potential problems) before they get out of control. Those managers would have to talk to hundreds of people every day to achieve the kind of high-level overview that a good data visualization provides. Talking to those on your organization’s front line is hugely beneficial, of course, but understanding the data will help managers know where to start those conversations.

Let’s take a very important operational task as our example. To understand how many people you need for each function, you’ll need to measure how long it takes to complete each task. If you don’t hire enough people, your team will get stressed and might not be able to complete the required tasks. And if calls aren’t answered or planes are messy, customers will quickly become unhappy. If you hire too many people, you might have happier customers, but your team won’t have enough to do and you’ll be paying too much in wages and salaries. Operations is all about that balance. So how can you use data to measure how long a given task will take?

Back in Chapter 5, you learned about distributions, like control charts, box plots, and whisker plots. These are really important techniques for sharing operational data. They use standard deviations to show not just the median or mean but also the expected values. That’s the data your operations team needs for planning. Claire’s control chart for them, shown in Figure 8-6, provide an overview of the data as well as allowing for closer inspection.

  Control chart of complaints at Prep Air with faded data points
Figure 8-6. - Control chart of complaints at Prep Air with faded data points

Claire’s control chart shows that typically the team can expect to receive up to 80 complaints per week for each department. The average number of complaints increased weekly until week 19 when the operations team had to take urgent action to address the ever increasing number of complaints. The operations team added more people to each of the three departments to ensure customers were receiving better service across their interactions with Prep Air. The Onboard team might have been particularly worried about week 19 but as the value sits outside the control limits, Claire should ignore that data point in her analysis as being an outlier.

For project managers, data is especially important: it allows them to measure progress, find blockages, and celebrate successes. Most organizations will only fund projects that include a clear timeframe for stages of progress and deliverables, and to set that timeframe, managers need data.

The source of data for operations departments is likely to be a project management system. Large organizations often use specialist project management systems to track progress and hold project data, whereas small organizations might simply keep it all in Excel. Either way, you can use that data to look at project overruns, predict that your team might have conflicting priorities on different projects, and plot out schedules, among other things.

Using data from Prep Air’s project management database, Claire creates the visualization in Figure 8-7, which looks at projects that are overrunning their estimated schedules. From these charts, we can see that some managers are delivering projects on the weekends-- a sign that their capacity might be stretched. Ideally projects would be completed during a weekday so people are not being forced to work at weekends if that isn’t normal behaviour.

  Project overrun dashboard
Figure 8-7. - Project overrun dashboard

Claire adds interactivity to this view, so her audience can focus on individual task owners or departments. The chart on the right of Figure 8-7 is a Gantt chart, which shows milestone dates for each stage of a project and estimates the time those stages will take to complete. Claire knows all too well that these time estimates are constantly changing, so she bui lds a dashboard using a data source that refreshes regularly. Updated information helps decision makers know whether they need to add additional resources to a project or investigate delays.

Marketing

To understand how she can help the marketing department, Claire turns to Alex, Prep Air’s head of digital marketing. “How would you say data affects marketing?” she asks.

Alex lets out a low whistle. “You could just ask how it doesn’t affect marketing. Honestly, the rise of digital marketing creating more data has transformed the whole field. It’s so much more crowded and busy. It’s harder to make your brand stand out. When I started out in this field, we used to have to rely on focus groups and surveys-- marketing research could really only happen one individual at a time. We still do some of that, but now we also collect and track data on social media engagement, web traffic, you name it. You can learn a lot about how customers perceive your brand before you even talk to any of them.”

“Where does all that data come from?”

“Good question,” Alex replies. “The biggest challenge you’ll face when working with us marketers is collecting and collating all those data sources. It’s a vast amount of information with different sources having different naming conventions and definitions to merge and clarify.”

Alex explains that Claire will need to pull together information from datasets like census records, web traffic reports from Google Analytics, and social media sites like Facebook, LinkedIn, Instagram, and Twitter to create a consistent view that links the most important messages from each. Tracking web traffic on specific campaigns is important too.

“When you click on an ad or a link in an email,” Alex elaborates, “you’ll see a specific URL, or web address, flash on your screen before you’re redirected to the product page. We can track which clicks came from which URLs, so we know which ad or email you clicked on. Now we know what got your attention and led you to our site. Then we can track your session on our site, so we can see how you navigate around and whether you buy anything.”

The ultimate goal is to link customers’ accounts to their actual purchasing behaviour. This not only allows marketers to understand who is buying what, it also lets them reverse the data flow to see what those customers are saying about the brand on social media. You can also find out if others with similar profiles might be interested in your product (or not). This lets you find the people who are most likely to want to purchase your products, so you get the most bang for your marketing buck.

To create those links, Claire will need to match up the fields from each dataset: for example, “Name” on a census form should be linked to “Name” on Twitter. Claire, an avid Twitter user, knows that this could get tricky fast: her government name is Claire, but the name she uses on Twitter is different as Claire was already taken. She quickly comes up with several other potential problems:

Facebook

Facebook uses real names, mostly, but someone’s Facebook name might leave out a middle name, include a former name to help old friends find them, or add a nickname.

LinkedIn

LinkedIn uses real names, too, but people who change their names when they marry often still use the old name at work.

Census records

Government census records include lots of details, but at the household level. It can be difficult to tie an individual to a specific household and to differentiate each adult in that household. In addition, people’s official government names don’t always match the names they use-- for example, transgender people sometimes have difficulty obtaining an official name change when they transition.

Web traffic

You might ask customers to enter their email address on your website, but if they don’t log in or you don’t hold their email address, it’s difficult to see what other websites they have visited.

These data sources are just the tip of the iceberg. You can see how linking them all together to form a profile of one identifiable person can be difficult to do. You will likely need to complete a lot of data cleaning to make the values you find from each of the data sources consistent with each other. Just identifying how to link the data sets together isn’t enough to form a perfect data set for your analysis. You will inevitably need to spend a lot of time forming the data set suitable for your analysis. Building profiles of customers and potential customers, though, is well worth the effort, because they allow you to target your marketing campaigns much more accurately.

For example, Prep Air would like to focus on individuals who fly frequently for business. This group spends frequently on plane tickets, so it’s a great market. Alex and the Prep Air team hope to identify those fliers’ preferred destinations by using the geographical data in their social media posts. To do that, they need Claire to link a dataset of customers enrolled in the Prep Air loyalty program with a dataset of their social media accounts. Alex would love to be able to know who the fliers are and where they go to add special rates on those flights to ensure they continue to fly with Prep Air and not potentially go for a competitor. Social media is a great way for organizations to see changing consumer behaviour as it happens so the Prep Air team might even notice new destinations becoming popular or emerging.

Linking datasets together to combine useful fields from each is called joining. To join two datasets, you need to specify two things: the join conditions and a join type.

Join conditions

Join conditions are the logic that specifies how two separate tables should be linked together. This is where you match one field with another: for example, customer_id in the first table matches up with customer_id in the second table.

Join type

Depending on the join conditions, you can choose whether to retrieve only records that match (this is called an inner join) or all records from each table in turn, regardless of whether there’s a clear match.

  Join logic
Figure 8-8. - Join logic

The resulting table of data depends on the type of join you create:

Inner Join: The only values returned from either table are where the join condition is met (Table 8-1). This is most useful when needing data from both tables to make meaningful analysis. Where the join condition is met, the data appears on the same row based on that condition

Table 8-1. Inner Join results
customer_id name loyal_prog
CI8932 Oscar Gold

Left Inner: All values are returned from the left hand table and data is added from the right hand table where the join condition is met (Table 8-2).

Table 8-2. Left Inner Join results
customer_id name loyal_prog
CI8932 Oscar Gold
CI2405 Janet

Full Join: All data is added from both tables but when the join condition is met, the data is added to the same row (Table 8-3).

Table 8-3. Full Join results
customer_id name loyal_prog
CI8932 Oscar Gold
CI2405 Janet
CI3496 Silver

The challenge, as Claire surmised, is lining up the information in all those different data sources to find common fields. Once she identifies those, she can create the join conditions she needs to form this full view.

Sales

“I love data!” says Michiko, one of the sales leads at Prep Air. “In sales, we’re really driven by targets. We have to make our sales quota if we want to make money. And I don’t know if I’m doing that unless I have data. I can’t wait to see your dashboard, Claire!”

Sales monitoring systems, also called customer relationship management (CRM) systems, contain a lot of data, but that doesn’t mean their data is easy to communicate. Many CRM systems offer built-in visualisations, but these usually can’t be customized to answer specific questions. “That’s where you and your data skills come in,” Michiko tells Claire. .

The sales team measures their success by the progress they’re making against their targets, but what they really need to know is why they are (or aren’t) hitting those targets. Depending on the product, a single large deal might be enough for a salesperson to make their annual target. So the sales department needs to understand the pipeline, or the list of potential deals and how they are progressing towards completion. The sales pipeline has multiple stages, from initial prospecting for clients to closing the deal (Figure 8-9).

  Diagram based on Salesforce s Sales pipeline stageshttps   www.salesforce.com ca hub sales what are the stages of a sales pipeline
Figure 8-9. - Diagram based on Salesforce’s Sales pipeline stages1

Not every opportunity moves successfully through every stage of the pipeline. The sales department wants to track the progress of each opportunity and identify any issues or blockages they need to address in order to complete the deal.

There are many ways to visualize the sales pipeline, but all of them need to communicate a few key pieces of information:

  • The value of the pipeline as a whole

  • The likelihood that each opportunity will convert, or create revenue with a sale

  • How long each opportunity takes to convert

  • Any changes in these measures compared to the previous period

  • How each salesperson’s pipeline compares to those of their peers

If Claire tried to build all of these elements into a single visualization, it would likely be too complex and difficult to understand. She decides that a dashboard is a better form to share this information.

  Dashboard showing the Prep Air sales pipeline
Figure 8-10. - Dashboard showing the Prep Air sales pipeline

Monitoring an account’s progress through the pipeline is a lot like tracking a journey. It involves understanding how long it takes for an account to progress to sale or rejection, and what happens along the way. Tracking trends in this data can generate a useful analysis of customer behaviour. Sales leaders need to know what efforts their team is making and how effective those efforts are.

For Prep Air, Michiko explains, “If our sales team are spending a lot of time focusing on landing large accounts and they neglect lots of other accounts, the business might suffer overall. We celebrate large sales, of course, but we have to balance that effort and resources to make sure we’re focusing on the right accounts.” To make this analysis, Claire will need records of all previous statuses of each account and when those statuses changed (Table 8-4).

Table 8-4. Useful data structure for sales data
Account Account Owner Product Type Estimated Val ue Status Date
PA-302818 Jenny Corporate 350000 Prospect 19/12/2020
PA-302818 Jenny Corporate 290000 Quote 23/01/2021
PA-302818 Jenny Corporate 290000 Invoice 08/02/2021
PA-302818 Jenny Corporate 290000 Purchased 21/12/2021
PA-193842 Tom SME 34000 Prospect 02/01/2021
PA-193842 Tom SME 34000 Quote 08/01/2021
PA-193842 Tom SME 34000 Invoice 11/01/2021
PA-127492 Tom SME 12900 Prospect 13/02/2021
PA-123428 Tom SME 12400 Prospect 17/02/2021
PA-387492 Jenny Corporate 125000 Prospect 19/02/2021
PA-387492 Jenny Corporate 140000 Quote 13/03/2021

This structure doesn’t always make it easy to measure the time elapsed between each stage, but it does capture the changing value of each deal. This helps predict the likely conversion rate of the deal and how much of the original estimate is converted. This in turn helps the business predict cash flow, set future targets and budget appropriately.

Information Technology

“You couldn’t do any of what you do without us,” Jamie, assistant to the CTO, tells Claire. “You wouldn’t have any datasets if it weren’t for the IT department. We built the systems this organization runs upon. We capture the data they produce. Without that, the executives would just have to rely on gut instinct and experience. You need us.”

Claire laughs at his tone, but she knows Jamie is right. Her dashboards have been so successful that department heads are now requesting monthly visualizations they can distribute to their teams. Building all these dashboards is too much work to repeat over and over: if she’s going to need to do this for every department every month, Claire needs to productionalize her data: that is, she needs to be able to produce it on a regular schedule without too much time and effort. She’s talking to Jamie because she knows that the IT department has procedures to carry out each part of the process and ensure that the data is robust. This includes identifying the sources of data, preparing a clean dataset, and producing the actual communication.

For example, Claire has been working from extracts of database queries, but Jamie has access to all of the organization’s databases. He also has the coding skills and powerful software to move that data from certified sources into a feed, which will make it much easier to update the datasets that power Claire’s communications. His process will look very different from Claire’s, and it will take time to develop but will make producing the analytics faster in the future.

Jamie will need to break down the logic Claire has used to form her datasets through filtering and calculations, then build it into the same software he uses to run similar productionalized work. The same is true for visualisations: using different software may change some of the aesthetics, but Jamie and Claire agree they’ll need to take care not to lose the key messages their audiences need to understand. They refer back to the questions Claire listed when she was generating requirements, to make sure the new version of the work still answers them.

“Now, once we productionalize the data communication process,” Jamie mentions, “you’re going to have a lot less control over how to iterate the work and adapt it to changing circumstances.”

“That sounds frustrating,” Claire replies, thinking of all the rapid iteration she did to create these communications.

“I know. But it’s important to make sure no one can just change the data or the visualisations on a whim. You’d have all sorts of mistakes sneaking in. Once we’ve certified that these communications are sourced correctly, you don’t want anyone messing around with them.”

“I understand that,” Claire says, “but I have to keep our communications relevant. I’m okay with having a controlled process, but I need to be able to iterate quickly when the CEO needs different information.”

“That’s a good point. Our team maintains thousands of reports, so it can be hard to get them to make changes quickly. Let’s meet a few times while we’re productionalizing this, to go over it and make sure you know how to make changes.”

The final part of the production process is knowing when to decommission your work. As the author, Claire knows the purpose of the work, so she’ll probably have a sense of when it isn’t relevant anymore. Since IT has limited resources, Jamie doesn’t want them working on data feeds or communications that are no longer required. This decision shouldn’t just rest on Claire’s shoulders, though: she’ll need to talk to her audience to see if they still find the work relevant and useful.

Summary

Working with different departments is an exciting opportunity, and one I’ve always enjoyed. When your data skills shine within your current role, you’ll likely be asked to interact more widely with departments and teams across your organization.

When you understand the needs and challenges of your colleagues in other departments, you’ll be prepared to give them what they need quickly and effectively. Remember that your colleagues are probably already working with data to some extent, at different levels of data literacy. You will need to be highly agile and fit your approach to each situation, just as Claire did. This chapter has shown you some of the challenges, terms, and requirements you are likely to encounter, but it is most definitely not an exhaustive list.

Knowing the unique challenges of each department will also inform your decisions about how best to help them, how best to structure and store their data, and how to spot the broader questions underlying each question they ask you.

The skills you have learnt throughout this book, if you use them well, can open new opportunities to you. When you build your knowledge of other departments and foster collaborative relationships with your colleagues there, you’ll find that all of your work develops and deepens in ways you might never have expected.

1 https://www.salesforce.com/ca/hub/sales/what-are-the-stages-of-a-sales-pipeline/

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
44.222.118.194