Top 5 Data Problems Small Businesses Face
The majority of small business owners provide specialized services or products that they are deeply passionate about. The owners want to share their craft with the community and prosper while doing so. Often, these business owners do not come from a technical background and “build the plane as they fly.” There is absolutely nothing wrong with this approach, as this mindset is likely the reason why they ended up as entrepreneurs and will, subsequently, succeed in business. That being said, there are a number of recurring problems that entrepreneurs face; areas where the team at Blu Sky Data can help them.
1. Data lives in silos
Sales data sits in the CRM, financials in QuickBooks, marketing in Google Analytics, operations in Excel, and customer support in yet another tool. Nobody can answer a simple question like "which marketing channel brings in our most profitable customers?" without manually stitching exports together. This is probably the single most common complaint. This manual reporting can take up days of someone’s month and results in data being backward-looking instead of real time.
This is often exacerbated by tool sprawl as an organization gets larger. One department might be using Salesforce for their CRM while another might still be using Hubspot. This causes constant disagreements and inability to reconcile between these departments.
It may be the least sexy part of the data ecosystem, but data engineering serves as the cornerstone of a business’s data strategy. Nothing tangible can be drawn from your disparate data sources without proper architecture. Marrying data sources will take an entrepreneur to the next level by being able to figure out what is working, what isn’t, and how to scale their business efficiently.
2. Tracking through spreadsheets
Spreadsheets are where most small businesses start…and many never leave. Honestly, there is always a time and place for a spreadsheet, even at a company as big as Amazon, but they should never serve as the source of truth for trusted business metrics. It may start as just tracking some inventory metrics in an Excel file, but it can balloon into many other metrics being tracked in numerous other spreadsheets. These problems will compound over time. There end up being multiple versions floating around, broken formulas, no audit trail, one person who understands the "master" file, and nothing that scales once the volume grows. This is tightly correlated with the disparate data sources above and solving the first problem generally solves tracking through spreadsheets as a source of truth.
Spreadsheets, and processes related to them, are often paired with knowledge trapped in people's heads. Institutional knowledge about how the data works, why a field is used a certain way, or how a report is built often lives with one employee. When they leave, the context leaves with them.
3. Poor data quality
Duplicate customer records, inconsistent naming ("Acme Inc." vs. "ACME"), missing fields, free-text entries where a dropdown should be, and stale contacts. Most owners underestimate how much of their analytics pain is really a data-entry and hygiene problem upstream. There is only so much reconciliation that can happen once the data is in the warehouse. For instance, “Acme Inc.” and “ACME” can likely be matched through fuzzy-text matching, but leveraging the customer’s email domain to infer the corporation on the front-end is the more elegant and scalable way to keep data hygiene up.
Even at large organizations, raw data can enter the warehouse very messily. I worked at an organization that was doing over $100M in revenue per year where we still faced issues you think would be solved for a company of that maturity. One common struggle was marrying company names from multiple types of data (internal products and CRM). These were all text input fields and email domain was generally the best solution to join them. Another struggle was figuring out what was our ideal customer profile, programmatically. We never asked customers for their job title because the product team was worried that it would detrimentally impact conversion rate (*cough* which should have been tested *cough*). In order to try to get job title onto customer records, we had to use software called Zoominfo, which was not always accurate and was certainly not complete (the majority of customer records were not enriched).
Do as much as you can on the front-end to standardize your data so your team is not wrangling it on the back-end, wasting valuable hours that they could be devoting elsewhere.
4. Not knowing which metrics matter
Owners are often swamped with dashboards full of vanity metrics (page views, followers) that come with native reporting from their software subscriptions. However, they lack the handful of metrics that drive decisions like customer acquisition cost, retention, gross margin by product, and pipeline conversion rates. Being able to establish these metrics are all the result of consolidating your data through pipelines and creating curated models to marry them for insights.
You need to ask yourself what the goal of your business is within the near term, medium term, and long term. Only then, will you be able to choose which metrics to optimize for. For example, if you are a new company focused on growing as fast as possible, then your primary metrics should be new customer and revenue growth. Secondary metrics should be customer acquisition cost (to monitor how well your money deployed in the market is working) and user retention (make sure you don’t just have “one and dones”). At this point, you should not be optimizing for your profit margin. You need to pay attention to it as a guardrail metric so the company does not go bankrupt, but growth is the focus.
Turning data into business decisions is another issue embedded within the metrics. Even with clean data and good dashboards, many owners struggle with interpreting what the numbers mean and deciding what to do about them. Insight without action is just reporting! Using our example from earlier, an owner might see that their new customer growth is doing great, but their retention is falling off a cliff after month 1. This can indicate that they are acquiring the wrong types of customers. Perhaps, they need to try out some promotions for subsequent purchases to create stickiness. Maybe they should have post-purchase surveys to understand if the product is satisfactory. This is something that Blu Sky Data can help you with to identify the root cause and get your business back on track for success.
5. Uncertainty about what to build or buy
Owners hear about data warehouses, business intelligence tools, and AI. How can they tell what's proportionate for a business their size? Most of the time they fear overspending on infrastructure they don't need. Occasionally, they’ll be concerned with underinvesting and hitting a wall later. It can be hard to determine what to purchase and when is the right time. There are loose guidelines on when you should adopt certain products or methodologies, and Blu Sky Data is here to help you determine if you’re at the right maturity for specific tooling.
It doesn’t make sense for a new company that is still struggling with data living in silos to be prioritizing building out a customer churn prediction ML model. However, this would make sense for a company that has curated data pipelines, tracks the right metrics (customer retention segmented by various characteristics), and is able to properly train this machine-learning model on data they understand.
While these are the top 5 problems that a small business might face as their data strategy matures, there are numerous other problems that will continue to plague companies no matter how large the size. The most common one that we encounter is data governance: how can we make sure that two teams are reporting the same value for revenue? New customers? Number of companies served? Data governance becomes increasingly important as a business grows, but it becomes impossible to achieve if the problems cited above are not dealt with first. Reach out to us at Blue Sky Data if you are facing any of these problems or need help with your data strategy, in general, and we would be happy to develop a plan to help your business achieve its goals.
