Category: Data Insights

3 Types of Insight from Data Analytics

Business owners who get dashboards without anyone interpreting them often tell me the same thing: they feel flooded. Revenue, job counts, close rates, cost per lead, margins by service line. Every chart looks like it’s asking for something, but not every number warrants a reaction. Treat every metric as urgent and you end up in a constant state of panic or stress. Treat everything as noise and you miss the opportunities sitting right in front of you.

So which numbers should you focus on, and which can you let go? The honest answer: it depends on the type of insight the number is giving you. Many analytics tools give businesses more data and information rather than actual insight. A dashboard can show you what’s happening – deciding what it means and what to do about it is a separate step.

From Numbers to Insight

In a previous article I covered the three pillars of effective data analytics. I won’t repeat them here but the short version is this: once we understand the context, the numbers, and the visuals behind our data, we can start pulling insight from it. That insight comes in one of three forms: contextual, directional, or actionable.

You may have heard of “actionable insight” as it has become one of those business buzzwords that gets overused to the point that it can mean almost anything (or nothing). Actionable insight is the most impactful of the three, but it isn’t the only kind, and mistaking the other two for it causes real problems. Getting insight matters – knowing how to respond to each type matters more.

If you’ve worked with data before or made decisions based on data, these three types line up loosely with how analysts would describe their work: first describing what happened, then diagnosing why, then prescribing what to do. My version keeps the focus on what owners need, which is how to respond.

Contextual Insight: Good to know, can’t change it

Contextual insight is something we can’t change but it sheds light on the rest of our data. It usually, though not always, comes from outside the business. Examples include rising material costs, interest rates, or a new competitor opening down the road. Seasonality belongs here too – seasonal industries experiencing a dip in revenue in the down months is expected.

The right response is to absorb it and adjust your assumptions. Update pricing, revisit budgets, and reset expectations for the quarter. Even though contextual insight doesn’t call for direct action, it’s still important. When a metric changes dramatically and we have no explanation for it, we haven’t finished analyzing the data. Contextual insight often supplies that explanation. That allows us to adjust where we can and move forward with confidence instead of guessing. A common mistake with contextual insight is trying to fix what you can’t control. You can’t change the cause of a contextual insight but you can change how you consistently respond to it.

Directional Insight: Wirth Watching, Not Yet Worth Acting On

Directional insight is good to know but isn’t mature enough to act on yet. It usually shows up in one of two ways:

• An anomaly. This is a one-off with a known explanation, such as one huge commercial job skewing the month or a week of rain. Note it and move on.
• An early trend. This is a metric that keeps moving in the same direction over time. This is the one to watch closely.

Spotting a trend early is where data gives you an edge, because you can make changes before you feel the full impact. Directional insight is essentially a warning sign saying: this might be something, or it might be nothing, so let’s watch it closely enough to find out which it is. A good example: average job size in one city has slipped for two months in a row. The right response is to watch, dig, and gather more data. Set the finish line up front with a threshold or timeframe that tells you when it’s time to act. For example: “if average job size in that city drops for a third straight month, or falls below $1,650, we take action.” That way you catch the snowball before it starts rolling, instead of debating it every month. 

As you can imagine, a common mistake here is either overreacting to a single data point, or ignoring the signal until it becomes a problem.

Actionable Insight: Clear Gap, Clear Owner, Clear Next Step

Actionable insight identifies an area where we can make a change and measure the results. It’s what makes data analytics powerful. Sometimes it starts as a directional insight that has been watched and confirmed. Sometimes it’s actionable insight the moment you see it. A customer being double billed doesn’t need three months of trending. Either way, actionable insight needs a clear gap and someone who can own the fix. Ideally, it has a clear cause too.

A quick example could be one crew’s revenue per week lags the others by 20%. Digging in shows why. Their jobs are scattered across the service area, so they spend about 40 more minutes a day driving than the other crews. That’s a clear gap with a clear cause we can act on this week. The right response to actionable insight is to decide, assign an owner, set a deadline, and measure the results. Tighten that crew’s routing, give it 30 days, and compare revenue per week again. If it works, roll it out to the other crews. Even crews without the same problem may benefit, which lifts revenue across the board. Start with a small change in one area, measure it, then scale what works. Be cautious of the common mistake of acknowledging the insight and never following through. You see the trend, you know the cause, you have a solution in mind, and nothing changes. This happens more often than you’d think.

How To Tell Which Type You’re Looking At:

These types of insight are not fixed. A contextual trend can create a directional signal, and a directional signal can become actionable once it’s confirmed. Here’s how that might play out. Material costs rise, which is contextual. A few months later margins on one service line start slipping while the others hold steady, which is directional. Once the pattern is confirmed, the cause becomes clear: that service uses more of the affected material, and its pricing never changed. Repricing that service is an actionable step. Following insight through that progression takes consistent attention over time. That’s hard to do while you’re also running the business. Hiring a data analyst or a team that can respond to each type of insight appropriately is where you can see significant change in your operations.

In every monthly or quarterly report for my clients, I label each finding as contextual, directional, or actionable, so they know at a glance what deserves attention. Contextual insight comes from two places: the industry knowledge clients bring, and the outside context I track that they don’t have time for. I show how it appears in the numbers and where it’s likely coming from. For directional insight, we define what would confirm or dismiss the signal before our next review, so nothing gets overreacted to or forgotten. For actionable insight, I give specific recommendations with expected impact, then track the results and adjust as needed.

Every analysis we deliver at Levo Insight must be truthful, valuable, and actionable. Part of being truthful is being clear about which findings call for action and which don’t. Pretending every number needs a reaction is exactly how owners end up flooded in the first place.

The doctor who washed his hands and lost his career

Ignaz Semmelweis. If you know this name, I applaud you. I first came across him in Brent Dykes’ book Effective Data Storytelling, and his story has stuck with me ever since, because it’s the clearest example I know of good data losing to poor communication.

Semmelweis wanted to know why. The key difference was that doctors and students started their mornings performing autopsies, then went straight to examining women in labor. Midwives didn’t conduct autopsies. When a colleague of his cut his finger during an autopsy and died of an illness that looked just like childbed fever, Semmelweis became convinced that “cadaverous particles” were traveling from the corpses to the mothers on the doctors’ hands – what we now call germs.

Keep in mind, this was decades before anyone understood germs. Gloves wouldn’t show up in operating rooms until the very late 1800s and handwashing between patients simply wasn’t a thing. So, in 1847, Semmelweis required everyone coming from the autopsy room to scrub their hands in a chlorinated lime solution. Ordinary soap wouldn’t get rid of the smell transferred to doctors’ hands from the corpses which Semmelweis used as proof the particles were still there.

The results were dramatic. Deaths in the first clinic fell from around 10% to about 2% – roughly a 90% drop. Something that drastic should have been celebrated as a breakthrough. Instead, it was largely rejected. Semmelweis wasn’t asking people to take his word for it, he had the numbers to back it up. He believed the numbers would simply speak for themselves – a trap some data analyst fall into – but they didn’t.

His handwashing practice was widely accepted long after his death. As Dykes points out, if Semmelweis had communicated his findings effectively, far fewer mothers would have died waiting for time to break the path for the truth – in Semmelweis’ own words.

Why weren’t the number enough?

Semmelweis’ colleagues are seen as stubborn but think about what accepting his data meant for them: it meant admitting that they, the doctors, were the ones killing their patients. No table of numbers was going to make that easy to swallow. It’s easy to criticize his peers but truth is we are all like this more than we’d like to admit.

Think about the scale in your bathroom. When it shows a lower number than you expected, you step off and go about your day with a smile. When it’s higher, you check that it’s level, step back on the scale, and shift your weight so there’s not too much on one leg over the other hoping the second reading reveals what you initially expected. Studies have shown that we are much less skeptical of information that tells us something we have already decided to be true or pleases us. Decisions are rarely made by logic alone. How information makes us feel plays a much bigger role than we tend to realize or admit. That doesn’t mean the data matters less, it just means how the data is delivered matters just as much.

With the benefit of hindsight, I see two things Semmelweis could have done differently: tell a story and show a picture.

First, the story. “A 90% drop in mortality” is a statistic. Behind that statistic were real women who came to the hospital to bring a life into our world but instead never went home. Putting faces on the numbers makes them much harder to dismiss. And the research backs this up. A 2022 analysis of 50 studies by Jie Xu found that stories and statistics are about equally persuasive overall, but stories are more effective at getting people to take preventative action, which is exactly what Semmelweis needed his colleagues to do.

Second, the picture. A single, simple chart showing the death rate in the first clinic before and after handwashing would have said more in a few seconds than hundreds of pages of tables. When a grid of numbers is all someone gets, it’s hard to see what’s going on, and easy to find a reason not to look.

What does this mean for your business?

At Levo Insight, every analysis we deliver must meet three criteria: it has to be truthful, valuable, and actionable. Truthful means the numbers are accurate and honestly presented. Valuable means it’s about something that actually affects your business. Actionable means you can do something with it. But as Semmelweis learned, meeting those three criteria isn’t enough if the message doesn’t land.

Here’s what that looks like in practice. Picture a plumbing company whose monthly report shows a callback rate of 6%. As a number in a table, that looks fine, and it’s easy to move past. But when we dig in, we find that one type of job, say water heater installs, accounts for nearly half of those callbacks, and each return trip costs a few hundred dollars in labor and fuel that nobody is billing for. Now the report isn’t “callbacks are 6%.” It’s “water heater installs are quietly costing you thousands a year, and here’s the chart that shows exactly where.” That’s a story an owner can act on this week, not something that surfaces at the end-of-year review.

Just like in Semmelweis’ case, the hardest findings are often the ones that point back at how things have always been done. That’s exactly why they need to be communicated with care. None of this is about spin. Storytelling doesn’t change the data. It makes sure the data gets heard. When the numbers are truthful, valuable, and actionable, and they’re delivered through clear visuals and a story people can connect to, business owners stop running on gut feeling and start making faster, more informed decisions. Over time, those decisions add up. Semmelweis had the right answer along with the data to support it and still couldn’t get his peers to entertain his solution. Our job as the data analyst is to make sure your data never has the same problem.

How to Set Your Business Up for Clean, Usable Data

Many business owners don’t know where to start when we talk about best practices. At the end of this read, you’ll have the knowledge and the resources to guide you through establishing best practices for data entry – leading to an easier transition when your team is ready for data analytics and dashboard build-outs. From what I’ve seen, those who lack the consistency and knowledge of how to collect data, where to collect it, and how to maintain it are the ones who write off data analytics as something out of reach for their business. I want to walk through the actual steps of getting you to a point of confidence and clarity with your data entry.

First, why does this matter? It’s fairly simple, but it holds plenty of weight. Every dashboard you’ll ever look at, every report you’ll ever read, and every decision you’ll ever make from your data starts here – at the moment someone types something into a form. If that entry is inconsistent, mislabeled, or missing, it doesn’t just sit there quietly. It gets pulled into your reports, baked into your numbers, and eventually shapes decisions built on information that was never right to begin with.

The Foundation

First, we have to give everything – and I mean everything – a unique identifier. Clients, crews, employees, jobs, quotes all get their own ID, consistent across every sheet. A client’s identifier is the same on the quote as it is on the job and the invoice. This is absolutely the most important piece to get right before preparing for an easy data integration process. Without IDs, your only way to match a quote to a job is by name, and names are unreliable. When working with data, “Michael,” “Mike,” “mike,” and “Mike ” are all different entries. Add a last name and the opportunity for error widens even more. An ID ties every record back to one canonical entity regardless of how the name was entered – it prevents duplicates, joins your data together, and future-proofs anomalies.

Here’s what that actually costs a business without it. Say a landscaping company believes it has 340 active clients. Without IDs, “Mike Johnson” might exist as three separate records – one from the quote, one from the job, one from the invoice – entered slightly differently each time. The real number of active clients is closer to 310. Every average built on top of that count – revenue per client, jobs per client, close rate – is quietly wrong, and nobody notices until the numbers stop making sense next to what the owner knows to be true on the ground.

Consistency

Uniform templates and forms allow for clear standards and a clear process for how to fill them out. Typically, in small businesses, the owner sits down and creates the exact templates and forms needed, in great detail. After creating them, actually filling them out – with any potential anomalies in mind – helps establish a foundational format and process. Then comes potentially the most important part: replicating that through training the rest of the team. Two employees may have different opinions of what a term means, so clarifying that up front matters. If one form says “location” and one employee interprets that as county while another uses city, it creates inconsistency that’s invisible until someone tries to use the data.

That kind of small gap compounds fast. A report meant to show “jobs by location” becomes unusable if half the team logged county and half logged city – the report isn’t wrong and loud, it quietly undercounts every location that got split across two different labels, and the owner has no way of knowing which numbers to trust.

The final part of consistency is audits. No matter how clear the templates are, how much training you do, or how much follow-up comes with it, there will still be mistakes. Audits help identify who, when, and how those anomalies happened. That allows the business to run targeted training – not one-size-fits-all, but specific to how that employee is getting it wrong. This is where standards and process make a lasting impression.

Timeliness

We tend to push off documentation. Every industry does this – I have yet to meet someone who enjoys stopping mid-work to reiterate what they just did. But documentation is how we measure growth, and timeliness is a huge part of getting it right. The closer the documentation happens to the actual event, the better. Ideally, a system or workflow assigns a client ID and quote ID and logs the quote the moment it’s sent. That’s not always feasible, but it can be approximated with process.

Delay has a real cost. A crew member who logs a job three days later is working from memory. A job that actually took five hours might get logged as four, or a material cost gets rounded because the receipt is buried in a truck somewhere. On its own, one estimate is a rounding error. Across dozens of jobs a month, that drift compounds into a labor-cost or margin number that looks precise but isn’t actually true. I’d avoid setting documentation deadlines any looser than end-of-shift, though I do see clients land on end-of-week as their standard. It opens the door to more error, but it’s still better than no deadline at all.

Access

Next, we want to limit who has access to input data. Role-based access usually works best here – the goal is to restrict data entry to specific people to reduce inconsistency. This gets overlooked because it feels counterintuitive: more people submitting data should mean more data. But the value comes from consistent data, not from the volume of it. Giving only managers access to a given process or template means less general training, more targeted training, and higher consistency across the board.

That said, don’t over-correct into a bottleneck. If only one person can submit a given type of record, that person becomes a single point of failure – vacations, turnover, or a busy week can stall data entry entirely. Two or three trained submitters per role is usually the right balance: tight enough to protect consistency, wide enough that the business doesn’t grind to a halt when one person is out.

Training

Templates and standards only hold up as long as the people using them understand why they exist. That’s why training can’t live only in the head of whoever’s been there the longest. When a new hire learns the process informally, from whoever happens to train them that week, the standard drifts a little more with every new employee. By the time you have five years of “however so-and-so explained it,” your data isn’t following one system anymore, it’s following five slightly different ones. Building the templates and the standard into actual onboarding (not just handed off as tribal knowledge) is what keeps the process intact as the team turns over.

Centralize

A cloud-based system isn’t strictly required, but I’d push back on calling it optional. A local spreadsheet on one laptop is one crash, one spilled coffee, or one accidental deletion away from losing months or years of records – records that can’t be recreated after the fact. Platforms with automatic backups protect against that outright, and they also make the data easier to organize and access as more people need to touch it. Backing up any document you keep should be a standing part of your routine.

Organization

Organizing clients by as much detail as possible pays off down the line – city, service type, frequency, lead source. Once the basics are solid, capturing adjacent data like material cost, labor cost, and subcontractor activity adds even more value for your team to work with later.

Automation

Once the process above is running consistently, automation is the next lever – but it only works on top of good habits, not instead of them. Automating a messy process just produces mistakes faster. What automation actually buys you is the removal of the manual step where most errors creep in: instead of someone retyping a quote into a spreadsheet, the form submission pulls straight into it. It’s more technical to set up, which is why it sits at the end of this list rather than the beginning but for a business with the foundation already in place, it’s often the single biggest time-saver available. I would caution from implementing automation without the foundational process. When the automation inevitably fails, it makes it difficult to troubleshoot without that prior knowledge.

The Payoff

None of this is exciting work, and it’s rarely the reason someone starts a business. But it’s the reason a dashboard build-out takes two weeks instead of two months of cleanup first. Every identifier assigned, every template followed, every record logged on time is one less thing that has to be untangled before your data can actually tell you something useful. Get this part right, and the transition to real analytics – the dashboards, the trend reports, the numbers you can actually act on – becomes the easy part.

If you like work from a checklist rather than re-read this, we’ve turned these steps into a free, downloadable guide (available under Resources on our website) that breaks each one down into measurable actions you can start on this week.

3 components to get the most out of your data

Data is a tricky tool – sometimes dangerous. It can have incredibly positive or negative effects on decision making if it’s not handled correctly. Clive Humby made the point back in 2006 that data is the new oil. Raw data has to be processed just like crude oil has to be refined before it has any value and oil that’s mishandled doesn’t just fail to help you, it causes a spill. Without the right tools and know-how, data can hurt your company just as easily as it can help it.

The small businesses I work with have the most to gain, and yet the most to lose, from trying to implement data on their own. Done right, data leads to understanding, understanding to action, and action to growth. But two mistakes we see constantly, no matter the size of the company, are 1. not understanding the numbers, or 2. taking the context of the business/industry and running with it – disregarding the numbers that back it up.

None of this is a new idea, and I won’t pretend I came up with it. Aswath Damodaran makes the case in Narrative and Numbers that a good valuation needs both a story and the numbers to back it up. Brent Dykes builds his entire framework in Effective Data Storytelling around three pieces – data, narrative, and visuals. Cole Nussbaumer Knaflic’s Storytelling with Data comes at it from the visual side: understand the context, choose the right chart, then tell the story. I’m sure there are plenty more. The framework isn’t mine, but it’s valuable, and what I want to share is how it applies to the businesses I work with every day.

There are three components necessary to get the most out of your data: Context, Numbers, and Visuals. As we walk through each one, you’ll see how each point of the triangle needs the other two to actually function effectively. 

1. Context

You may have heard of context called story, narrative, or external and internal factors. All the same thing here. Context is what’s going on in and around your business and your industry that provides insight. It generally comes from three periods:

• Historical: what has happened before – what solutions worked, what failed, and the circumstances surrounding it.  
• Current: immediate awareness of the environment and conditions you’re operating in right now.
• Future: with current and historical context together, we can anticipate what’s coming.

Let’s take this a step further. The context that matters most is context specific to your business and your industry – and the story you believe about it shapes every decision that follows. Take Uber. Back in 2010, Uber raised $1.25 million in its seed round. To get there, it had to convince investors to buy into a narrative. Investors either saw Uber as trying to break into the red ocean of the transportation industry – an already crowded market full of taxis, buses, and limos – or as one of the first movers into a brand-new ride-sharing category. Same company, different story. The ones who passed missed out on millions, even billions (just ask Mark Cuban, who turned down a $250,000 stake). Damodaran uses Uber as a centerpiece in his book for the same reason: in 2014 he valued it at around $6 billion, venture capitalist Bill Gurley argued the market was far bigger than Damodaran’s story allowed, and the gap between their valuations came down to the narrative, not the math.

Bring that back to trade businesses. A landscaping company in Ohio has different context than one in Texas – the seasons differ, the fertilization schedule differs, and so on. I think most would agree that knowing how terminating a service line, or letting a manager go, could impact total revenue is valuable context. Without it, a dip in revenue from one month to the next can psych you into taking action you don’t actually need to take. Context explains the numbers, and vice versa.

2. Numbers

I hear a myth that carries a lot of weight in business: “the numbers don’t lie.” I’d argue they certainly can and will if you don’t confirm they’re reliable. Correlation and regression can tell you how strongly one set of numbers moves with another, but neither can tell you why. You’ve probably heard this as “correlation does not equal causation.”

Back in college, I ran a small experiment on exactly this. I set out to see if household income could explain birth defect rates – something that, on the surface, should have almost nothing to do with the other. What I found: the lower the household income, the higher the rate of birth defects. I wrote it up like a genuinely interesting discovery, and I intentionally saved the regression results for the very end of the report.

Here’s why: correlation tells you whether two things move together. R² tells you how much of the change in one your model explains. Beta tells you how much y moves when x moves. None of them can tell you why. Income and birth defects were correlated, and on its own the regression backed that up. The statement “lower income causes birth defects” could technically look “true” if you only trusted the numbers on the surface. But once I added smoking rates to the model, income’s effect nearly disappeared. Income wasn’t the direct driver – it was standing in for something else.

The missing explanation? Lower income correlates with high stress, high stress correlates with higher smoking rates, and higher smoking rates during pregnancy correlate with birth defects. I left that fact out on purpose, because it explains the numbers far better than income alone does. Yet without proper context, you could convince someone that lower income directly causes birth defects. The same trap shows up in trade businesses: revenue jumps in April, the same month you start running ads, and it’s tempting to credit the ads – until you remember April kicks off your busy season every single year.

Now I don’t run a linear regression on every number that crosses my desk at Levo Insight, but we do double- and triple-check our numbers, because we know they can lie. One tell that your numbers might be lying to you: check whether they can be explained by context outside the numbers themselves. If they can’t, that’s usually when a deeper statistical check is the right next move. Numbers, paired with context, explain what’s actually happening.

3. Visuals

Most people think visuals tell the whole story on their own. That couldn’t be further from the truth. Visuals are interpretations of the numbers – nothing more, nothing less. As humans, we understand pictures fast. Our brains process shapes, faces, colors, and patterns almost instantly, which means there’s a lot less mental translation happening when you’re looking at a visual, and a good one can carry a surprising amount of context on its own.

Reading a spreadsheet and trying to spot where numbers moved over time, and how much, is tedious work. Seeing a line gradually climb to the top right of the screen tells the same story but understood in seconds instead of minutes or hours. In fact I’d bet you imagined that graph in your mind while you read it. If numbers make you think, then visuals let you see.

Visuals lead to insight, and the wrong insight leads to the wrong action. That’s why it’s important to choose the right visual for the job, even if that means a visually repetitive dashboard. An effective dashboard is worth more than a pretty one with new and crazy graphs. When you’re looking at something that adds up to 100%, pie charts, stacked bar charts, and area charts work best because the shape itself visually represents one whole, with each piece as a share of it. On the other hand, using a pie chart to show revenue over time would be a poor match – a line graph handles that far better. I’ll admit I reach for a fancier-looking chart now and then, but more often than not, the simpler the graph, the better it communicates, and the faster we can act on it.

How This Connects

Have you ever spent hours, or even days, studying a topic or trend, and then had to explain it to someone in seconds? You’ve built up detail after detail – layering context on top of numbers on top of visuals – and trying to hand all of that to someone starting from zero, in seconds, is impossible. MIT researchers found the brain can identify an image it sees for as little as 13 milliseconds – far faster than I could say the first word of my explanation.

If I tried to walk you through how revenue increased over time, except for an odd dip in month seven, which is strange because we actually added a new crew that month, but on the other hand we cut a service line with profit margins below 10%  but there always was a lot of crazy weather that month, and blah blah blah – I mean I’m snoring already. But if I handed you a dashboard on revenue alongside a plain English report explaining why we cut that barely profitable service line, and how a marketing adjustment drove a jump in revenue the following months, you’d take it in, understand why, and potentially act on it far more easily. Especially when presented through a story.

We covered three main elements: context, numbers, and visuals. When they’re intertwined, the relationships between them look like this: context and numbers explain each other, numbers and visuals help you understand each other, and context and visuals allow you to interact with each other.

As a business owner, it’s important to know your numbers, and even more important to understand them. Context, numbers, and visuals together give you that opportunity – and with it, the ability to make confident, informed decisions that move your goals forward. Decisions that improve efficiency and drive growth. The same decisions that turn a small business into an enterprise.

If you’re sitting on data you’re not sure how to read – or you’re not sure your numbers are telling the truth – that’s exactly what we do at Levo Insight. We build dashboards for home-service businesses and pair them with plain-English explanations of what’s driving the numbers, so you can act with confidence. Reach out at levoinsight.com, and let’s take a look at yours.