- Company profile
Formed in 1826, the Zoological Society of London is an organization that has dedicated its efforts to scientific activities, education, and protection of endangered species. The ZLS is a non-profit organization and depends on donations and gate fees for funds. The ZLS has two zoos in London which it manages. These zoos are ZLS London Zoo and ZLS Whipsnade Zoo. The ZLS educates people on the animals in the zoos and protects some wild animals which are on the blink of extinction.
The number of visits the park receives has recently been on the decline. A decline in the number of visits to the park indicates that the revenue required to run the park decrease. Generally, the number of visits to zoos has been on a decrease. This has been brought about by different happenings such as the spread of the Coronavirus. Previously the numbers were still on a decline compared to the number of visits in the years before 2016. There has been news on the possible total closure of the zoos. This is due to the lack of funds to cater for the needs of the animals. ZLS has employed the services of different professionals to help in the successful running of the company. These professionals include veterinarians, scientists, researchers, and wardens.
To overcome these hurdles, the implementation of a big data strategy is highly recommendable. Big data solutions imply that decisions no longer have to be made depending on a gut feeling. Decisions are now made based on information provided by data. Companies that have adopted the policy of making decisions based on data are more likely to succeed than those who depend on their gut feeling. For this purpose this report is written to provide the guideline and steps to implementing a successful data strategy.
Data Requirements
As stated earlier the main aim of implementing the big data strategy is to provide data-driven recommendations that will help arrest the decline that has happened and is threatening to lead to the closing down of ZSL. There is therefore a need to increase the number of visitors to the zoos to help in the continued facilitation of conservation efforts. Through analysis formulation of a new package for the visitors will be possible and the determination of competitive prices will be possible.
According to Schmarzo (2013), there exists some big data supporting requirements that assist in the gathering of data and its subsequent analysis.
Key performance indicators
The performance indicator exists as a tool for measuring the impact of implementing the big data strategy that will be proposed from data analysis. The performance measures will be the impact of new advertising strategies i.e., whether the visits from the advertising will increase, the impact on revenue, and generally the impact on the total number of people coming to visit per day. This will be possible through the analysis of data about previous visitors to the zoo. This data is collected over time and compared with past results that help gauge whether the new strategy was a success or a failure.
Business Questions
Business questions are formulated through analyzing the problem at hand and thinking about what should be done and the source of the problem. Business questions for ZSL include: For the number of visits to the zoo to increase what should be done? How can revenue be increased? Are the prices being offered by ZSL the best prices in the market? Which is the most effective channel for advertising? What are peoples’ views on different animals in the environment? Which specific group needs to be targeted to increase the number of visits?
Business Decisions
Business decisions are identified to:
- Conduct case summaries on different categories of the visitors e.g., members and non-members, students, educators, locals, and foreigners.
- To collect and analyze data from a preference survey.
- To determine the best price
- To use correlation to test for the relationship between the number of visits and the supposed dependent variables which include family size and salary among others.
- Develop a regression model to help forecast the number of future visitors.
The first decision will help determine the characteristics of the visitors and point to the various categories that are contributing less. This will enable choose areas that ZSL can choose to improve to achieve an increase in visits. The second decision is developed to help determine which products are best preferred. ZSL can, therefore, decide to either dedicate their efforts to improving products that are not preferred or drop them in favor of the highly preferred products. The third decision will help assess the impact of adopting a new piecing strategy. The best price will also be compared with other prices from other competitors to determine whether it is favorable. Testing the relationship between the number of visits and various characteristics such as where the ticket was booked, the profession, and how long they have been visiting. The fifth decision will help predict the number of visits to the zoo in 2019.
Determining and modeling the best analytic algorithms
It is important to know what any data set entails. This necessitates the use of descriptive summary statistics. The descriptive statistics will include frequencies, the means, and the case summaries. Case summaries that are selected for this strategy include case summary based on the number of family members, where tickets were booked, the purpose of visit, and membership.
Correlation analysis will also be applied to the data. The correlation will help determine the relationship between the salary and the average spending at the zoo, the age and the number of visits, number of visits, and average spending. Through the determination of the relationship between different variables, it is possible to determine which variable affects the other. Correlation also indicates a linear relationship between the variables.
The establishment of the presence of a linear relationship facilitates the building of both a linear regression model and a multiple regression model. The development of these models will help in the forecasting of the number of visits per year that a person is likely to make. The simple linear regression model can be developed for the total number of visits to determine the number of visits in the future.
Supporting data sources
The final step according to Schmarzo’s steps is to identify the data sources that will support the adoption of the data strategy. Historical data for the past 8 years has also been collected to show the trend of the visits. Data has been gathered from the history of previous visits to the zoo. Data has also been collected on the preference of the products that are found in the zoo. Reactions from ZLS’s twitter accounts are also collected for analysis. The data contains information about visitors and different characteristics which they exhibit. Prices form the price determination model are collected to help come up with a new pricing strategy.
Data Analysis
Descriptive statistics
The table below shows the descriptive statistics for the overall number of visits from 2010-2018.
| Statistic | Number of visitor(‘000) | Average number of visitors per month |
| Mean | 1155.89 | 96.32 |
| Median | 1134 | 3.436 |
| Maximum | 1319 | 109.916 |
| Minimum | 974 | 81.17 |
| variance | 7.5 | 106.25 |
The scatter plot below shows the trend for the number of visitors that the zoo receives.
The highest number of visits was reached in 2014 while the least was reached in 2012. However, in the years after 2016, the number of visits decreases and is expected to decrease further due to the effects that will come as a result of the Corona pandemic.
The table below shows a summary of the characteristics of the visitors.
| Age | Salary | Family member | Number of visits per year | |
| Mean | 35.30769 | 31.41026 | 2.794872 | 3.153846 |
| Standard Error | 1.780519 | 1.751698 | 0.208392 | 0.250609 |
| Median | 39 | 35 | 3 | 3 |
| Mode | 45 | 18 | 2 | 4 |
| Variance | 123.6397 | 119.6694 | 1.693657 | 2.449393 |
| Minimum | 18 | 14 | 1 | 1 |
| Maximum | 53 | 49 | 6 | 7 |
| Sum | 1377 | 1225 | 109 | 123 |
| Count | 39 | 39 | 39 | 39 |
The total number of participants whose characteristics will be analyzed is 39. The visitors have a median value of. This implies that there is an uneven distribution of the ages in the sample. This is further indicated by the value of the model which is 45. The minimum age of the visitors is 18 while the maximum is 53.
The figure below shows the bar graph for price analysis between three seasons.
ZSL currently employs a pricing strategy that is highly competitive in the market. The prices vary with those of competitors and in the general market with a small degree. Using the current prices, the highest count of visits is reached during the standard season. The least is made during the peak seasons. After the increase in prices, the count of visits is generally expected to fall as shown in the graph above. However the increase in prices is minimal and does not exceed the market price. Under this new piecing policy, the company would still offer very competitive prices.
Campaigns have been ongoing for the adoption of animals. The campaigns were carried out through different media i.e., print media, online, TV, and radio stations. ZSL uses different channels in the media for campaign activities.
The bar charts below show the cost versus the number of people who adopt animals as a result of the advertisement from a specific channel in print media.
For print media, the National Geographic channel has the best return on the investment of advertising. The channel yields a relatively high amount of revenue considering the cost and readership. The worst performing is the Sunday Times. The cost of doing campaigns on Sunday times is expensive although it yields an adoption rate of almost 100%. The revenues from campaigns on Sunday times are also relatively high. BBC wildlife magazine charges the least amount in advertising but also has the lowest readership.
The bar charts below show the cost versus the number of people who adopt animals as a result of the advertisement from a specific channel on TV.
From the bar graph above, the most expensive TV station to advertise on is ITV. However, this channel has the highest readership and also yield the highest revenue. Sky Television is the cheapest option but it also yields the least revenue. The effectiveness of the campaign on ITV is relatively high.
The bar charts below show the cost versus the number of people who adopt animals as a result of the advertisement from a specific channel on the radio.
The highest revenues are recorded from advertisements on Smooth FM. However the readership is highest in Heart FM which is also the most expensive channel. Heart FM also yields the smallest count in the adoption numbers. Advertising on Magic FM yields fairly good results because of its average cost and the high number of adoptions resulting.
The bar charts below show the cost versus the number of people who adopt animals as a result of the advertisement from a specific channel on the internet.
Through technological advancement, most people access online channels such as youtube which act as entertainment spots. However these channels also act as advertisement channels. Facebook and YouTube are among the most popular online advertising channels. This two yield the highest number of views among all other online channels. However, the number of clicks is greater on Facebook than on YouTube. Advertising on these two channels is also the most expensive. The online channel that records the highest number of unit sales is the animal fact guide followed by the wildlife article. The animal guide also earns ZSL the highest amount of revenues.
The analysis of people’s reactions on social media can be used to gauge the image of ZSL. The bar graph below shows the distribution of responses.
The online presence of ZLS is minimal. This implies that they are lagging on the adoption of social media as a tool for advertisements. The variations between the amounts of retweets display a high level of volatility.
Analysis of the preference products is also very crucial. This analysis will allow ZLS to note the product which is rated lowest and that which is rated highest.
The table below shows the analysis for the ratings of the products.
The highest-rated products are products that are labeled pref with an index above 20. In this range the lowest preference accorded to any of the products is 21. For the range between index 1 and index 20, a majority received a minimum rating of 1. Product 29 is the most preferred product with a mean preference of 27.00 while the least preferred is indexed 3.
Case summaries
The table below shows the case summary for the purpose of the visit.
| Purpose of visit | ||
| total | %frequency | |
| Exhibition | 11 | 28.2 |
| Library | 11 | 28.2 |
| Zoo experience | 17 | 43.6 |
The zoo offers a category of three services which are exhibition, zoo experience, and library. Visitors coming to the zoo can choose any of the services they require. This implies that their preferences differ. The highest number of people who visit the zoo is there for the zoo experience. This majority is represented by 43.6% of the total sample that was selected. Exhibition and Library has an equal share of visitors with 28.2% each.
The bar graph below shows the total contribution of the three activities that are mentioned above.
From the bar graph, zoo experience is the product that earns the Zoo the highest amount of funds. The least amount of funds are collected from students. Exhibition performs relatively average when compared to the total spending for the library and zoo experience.
Correlation
Correlation analysis helps in the establishment of the presence of any relationship between variables. The relationship between the number of visits to the zoo and the amount of money spent is -0.125. The negative correlation implies that as the number of times a visitor comes to the zoos increase, the amount of money he/she spends at the zoo is likely to decrease. The relationship between the count of time a visitor has visited and the amount of money he/she is likely to spend is relatively weak.
The table below shows the results for the correlation between the number of visits and the average spending.
| Number of visits /per year | Avg spending per visit (£) | |
| Number of visits /per year | 1 | |
| Avg spending per visit (£) | -0.12534848 | 1 |
The number of family members one has is also likely to influence the amount of money he/she is likely to spend when visiting the zoo. Generally, as the number of family members increases, the average spending is also supposed to increase. The value of the correlation between the number of family members and the average spending per visit at the zoo is 0.674. This is clear evidence that as the number of family members increases so does the spending.
The table below shows the results of the correlation test
| No. of family member | Avg spending per visit (£) | |
| No. of family member | 1 | |
| Avg spending per visit (£) | 0.674309287 | 1 |
Regression
From observing the scatter plot for the number of visitors, a linear relationship is observed. From this linear relationship it is possible to predict the future number of visitors by developing a simple regression forecasting model.
The table below shows a summary of the regression model.
| Regression Statistics | |
| Multiple R | 0.359405 |
| R Square | 0.129172 |
| Adjusted R Square | 0.004768 |
| Standard Error | 123.4004 |
| Observations | 9 |
The model has an r squared value of 0.3594 which implies that the model fits about 40% of the data.
| Coefficients | Standard Error | t Stat | P-value | |
| Intercept | -31538 | 32084.91 | -0.98296 | 0.358375 |
| Year | 16.23333 | 15.93092 | 1.018983 | 0.34213 |
The table below shows the details of the regression model. The p-value of the model is 0.35 which is above 0.05 alpha level of confidence. This implies that the values of the number of visitors vary with time. The equation of the model is:
From the model the prediction of future values for 2019 and 2020 is possible. From the model, visits for 2019 and 2020 are expected to be 1237.1 and 1253.333. Observing the two values leads can imply that the number of visitors increases but with a relatively small margin.
A multiple regression equation can be modeled to explain the relationship between the average amount that a visitor spends in a year as the dependent variable and other variables. The other independent variables are the visitor’s age, salary, count of members in the family, and the duration that they have been visiting.
The table below shows an analysis of the multiple regression model:
| Regression Statistics | |
| Multiple R | 0.988806 |
| R Square | 0.977737 |
| Adjusted R Square | 0.974364 |
| Standard Error | 10.23909 |
| Observations | 39 |
The multiple regression performs excellently. The r squared value of the model is 0.9777. This implies that this model fits about 98% of the data. The model is very accurate.
The table below shows the ANOVA to determine the significance of the model
| ANOVA | |||||
| df | SS | MS | F | Significance F | |
| Regression | 5 | 151943 | 30388.6 | 289.8596 | 2.96E-26 |
| Residual | 33 | 3459.688 | 104.839 | ||
| Total | 38 | 155402.7 |
The p-value from the ANOVA table is lower than the 5% alpha level of significance. This implies that the test is significant.
Recommendation
From forecasts of the linear regression model, the first recommendation is that for the survival of ZSL the amount of revenue must increase. As stated earlier ZSL is a non-profit organization and mainly depends on the fees collected during visitors’ entry and donations from well-wishers. From the regression forecasts, there is a high probability that the number of visitors will increase but at a very low rate. This low rate is not enough to sustain the zoos.
From the analysis of the visitors by profession it is recommended that education be offered on products and activities at the zoo. At present students exclusively visit the zoo for library purposes, while educators also mainly visit the zoo to access the library. These two groups combined to produce the highest number of visits for a product whose return is very low. Return from library services is relatively low, Education on other activities at the zoo to students, educators, and other local people is very important. This will lead to an increase in revenues as the number of visits increases. If there are no activities that can attract these three categories of people to the zoo, it is highly recommended that ZLS diversify its activities to attract this category. Tapping into local customers is highly recommended considering the recent outbreak of coronavirus which has led to the shutting down of international flights. This implies ZLS’s main source of income has been cut off.
The third recommendation is on the best channel to use for advertisement. Combining the three different media in advertising is highly effective. However, the choice of these channels should be in such a way that the money spent yields a high income. This is the cost-benefit analysis. ZSL should look to eliminate certain channels from those selected to advertise. These eliminated channels are channels that have a very high cost but yield low revenue. For the print media, the recommended channel for use is the Sunday Times. For all print media, revenue exceeds the cost. For the online channel, it is recommended the wildlife article be used as the main online channel for campaigns. ZSL should seek to eliminate Facebook and YouTube from its advertisement channels. For these two channels the cost exceeds the revenue. In the radio station advertisement option, the highly recommended channel is Smooth FM. This channel has a relatively high readership and yields far better results in adoption. It is highly recommended that ZSL adopt Smooth FM, Wildlife Article, and Sunday Times as their major advertisement channels.
It is also highly recommended that activities for large families be incorporated into products at the zoos. Of all the visitors selected for analysis, a majority had come to the zoo with family members. Their relationship between the number of family members and the average amount that is spent in a day is by a visitor indicates that if ZSL were to increase visits by large families then their revenues would also increase.
The scrapping of the products which are labeled pref with an index of less than 20 is highly recommended. These are the products that yield the lowest ratings among the 30 products. However, if the products serve a special category of the population, it is disadvantageous to scrap them. In case the option of elimination is not viable, these are the categories of products that ZSL should strive to improve. If the option of scrapping the first twenty products is viable, it is recommended that efforts go to ensuring the quality of the remaining ten products does not go down.