Getting the right blend
Michael Wagstaff • 29 September 2023
Merging survey data with business intelligence
In the modern era of data-driven decision making, businesses have access to a wide variety of data sources. While each type of data offers its own unique insights, the magic often lies in the blending of various data types.
One such powerful combination is the amalgamation of survey data with a brand's existing metrics like sales, revenue, CRM and customer experience data. This blending can provide an all round view of brand performance and help drive strategic actions.
Why Blend Survey Data with Existing Brand Metrics?
Survey data can tell you how your customers feel, while sales and revenue data can indicate how well the brand is doing financially.
Combining these can help pinpoint areas where customer sentiments and sales diverge, providing key areas of focus. In summary, the additional insights achieved by blending survey data with your own business metrics include:
- Granular Customer Segmentation: CRM and customer data can help you understand who your customers are. When this is layered with survey responses, you can gain insights into the preferences, pain points and sentiments of specific customer segments.
- Predictive Analytics: Combining behavioural data with attitudinal data (from surveys) can help brands predict future trends, customer behaviours and even potential risks or opportunities in the market.
- Mapping Customer Satisfaction to Sales Data: Imagine you've just launched a new product. While sales data might show robust numbers, your customer satisfaction surveys could reveal that many customers aren't happy with a specific feature. By blending the data, you identify a potential risk early on, allowing for timely product tweaks.
- Lifetime Value and Feedback Correlation: By integrating CRM data with survey feedback, brands can identify if their highest-value customers are the most satisfied. If not, targeted campaigns or personalised offers can be designed to enhance their experience.
- Tailored Marketing Campaigns: Let’s assume a segment of your customers, identified from CRM data, consistently purchases a particular product. If survey data shows this group desires more sustainable options, the brand can launch a marketing campaign highlighting the eco-friendly attributes of the product.
- Churn Prediction: If revenue data shows a decline in repeat purchases and survey data highlights growing dissatisfaction with customer service, these combined insights can lead to proactive interventions to reduce churn.
Blending the data
It's much easier to blend data sets that I have unifying variable such as email address, customer number or postcode. But unfortunately that is not always the case.
When two disparate data sets lack a common linking variable, integrating these sources is inherently challenging but not impossible. Strategic approaches are essential to ensure that blending such data yields meaningful, actionable insights.
Below are several methodologies and strategies to consider.
1. Aggregation & Segmentation:
Aggregate both the survey and internal business data based on common dimensions such as demographics or geographical location, creating segments that are comparable across both datasets. This can involve analysing trends, patterns and average values within specific segments rather than individual level data.
2. Propensity Score Matching:
Propensity Score Matching (PSM) is a statistical technique that can be employed to make inferences in the absence of a unifying variable. It involves creating comparable groups based on observed characteristics, allowing insights to be drawn by comparing the propensity scores of the survey data to the business data.
3. Canonical Correlation Analysis:
Canonical Correlation Analysis (CCA) can be used to understand the relationships between multivariate datasets when there is no common variable. This statistical method helps in identifying patterns and correlations between sets of variables in different datasets, thereby facilitating the combination of insights derived from both data sources.
4. Data Fusion & Feature Engineering:
Data fusion techniques involve creating new variables or features that are common to both datasets, facilitating more robust analysis. Feature engineering can include creating bins or categories that can be compared or finding proxy variables that can serve as a link between datasets.
5. Advanced Machine Learning Techniques:
Leverage advanced machine learning models that can handle heterogeneous data types and are capable of finding complex, non-linear relationships between datasets, even in the absence of a clear linking variable. Techniques such as clustering and principal component analysis can help in uncovering latent structures and relationships in the data.
6. Expert Judgment & Domain Knowledge:
Consult with domain experts who have in-depth knowledge of the subject matter and industry to formulate assumptions and hypotheses about potential relationships between the datasets. Expert input can help in contextualizing the findings and prioritizing which correlations or patterns are most meaningful and actionable.
7. Use of Proxy Variables:
Identifying proxy variables that are available in both datasets is crucial. While they may not directly correspond, they can offer an indirect link. For instance, socioeconomic status could be a proxy variable linking survey responses to purchasing patterns.
8. Trend Analysis:
Compare the trends observed in both datasets over time. While it won’t offer a direct link, it could provide valuable insights into how changes in one dataset correspond with shifts in the other, allowing for deeper understanding of the dynamics between customer feedback and purchasing behavior.
9. Qualitative Analysis:
Implementing qualitative analysis methods such as thematic analysis on open-ended survey responses can also provide insights that can be related back to existing business data, enabling the identification of overarching themes or patterns that can guide business strategy.
Our expert team is able to apply these methods to your datasets. Contact us for details.
Conclusion
Blending disparate data sources without a unifying variable is a complex endeavour but one that can yield significant insights when approached strategically. Employing advanced statistical, machine learning methods, and qualitative analysis coupled with expert judgment can unearth correlations, patterns, and trends between the datasets.
The integration of these diversified approaches ensures that even without a direct link, a harmonious synthesis of data is achievable, empowering brands with enriched, multi-faceted insights to drive informed decision-making and strategy optimisation.
The huge volume of data available through consumer comments, reviews and surveys can make cutting through the noise difficult. In this article we discuss how text analytics combined with human expertise can uncover the insight.
Market research agencies are going all in on AI based models to generate next level consumer insight. But are these just more illusion than substance?
With the online survey already on the ropes due to poor quality, has data science finished it off?
Research agencies are pinning their futures on AI. Are they right to do so or are we missing trick by ditching the human?
Online surveys suffer from fake, ill considered and unrepresentative responses. What can be done to improve their reliability? Triangulation is the key.
With so many agency panels riddled with fake respondents resulting in poor quality data, are we witnessing the end of the online survey?
With the January transfer window closed, we run our predictive model to work out the probabilities of where each team will finish in the final Premier League table.
In this latest article in our series on the power of text analytics we look at how sentiment analysis can be used to really understand what customers think about product offerings.
The true value of review sites lies in going beyond the stars and analysing what reviewers are actually saying.
Start making sense of customer review and feedback data by using text analytics. In the first of a series of articles we discuss how it can help your business.