Analyzing Website Traffic and Social
Media Engagement to Refine Content
Strategy for The Henry Ford
OVERVIEW
CONCEPT
This Digital Analytics project helps The Henry Ford analyzing their website traffic and social media engagement to refine content strategy.
INTRODUCTION
The Henry Ford is a not-for-profit organization that aims to inspire people to learn from America's traditions of ingenuity, resourcefulness and innovation to help shape a better future. This Web & Social Media Analytics project helps The Henry Ford identify its traffic sources to the website and gives an in-depth review of the social media content posted on the 3 platforms - Facebook, Instagram and Twitter. It also analyzes the museum’s social media activity and its impact on driving traffic to the website and provides recommendations based on the findings to improve user engagement.
DURATION
8 weeks,
Teamwork
SKILLS
Web Analytics
Social Media Analytics,
Data Cleaning,
Data Analysis, Dashboard,
Mockup,
Visual Design
MY ROLE
Data Analyst
Data Researcher
Visual Designer
TOOLS
Google Analytics,
Google Sheets,
Sprout Social,
Looker Studio,
Figma,
Piktochart
ABOUT THE CLIENT
The Henry Ford is a not-for-profit organization that aims to inspire people to learn from America's traditions of ingenuity, resourcefulness and innovation to help shape a better future. The Henry Ford in Dearborn, Michigan houses a collection of artifacts from 300 years of American history. It has a total of over 1.7 million visitors annually at all of its venues combined. The venues include - Henry Ford Museum of American Innovation, Greenfield Village, Ford Rouge Factory Tour, Giant Screen Theater and the Benson Ford Research Center.
The collection of artifacts mentioned above is also available on the museum’s website. The museum wants to promote this digital collection through the different social media platforms available and increase traffic to the website. The focus of this project was to help the Henry Ford Museum identify the social media platforms that brought traffic to their website and understand user engagement on the social media platforms(Facebook, Instagram and Twitter) to help promote these digital artifacts.
Google analytics data helped identify the traffic sources (all + social media) to the website, identify most visited landing pages, and analyze campaign performances. The sprout social data provided by the museum helped carry out a quantitative analysis to identify the top performing content types and categories for Facebook, Instagram and Twitter posts. A qualitative analysis of the posts on these platforms helped determine areas of improvement for the post content.
ABOUT THE TEAM
This Web & Social Media Analytics project is a collaborative effort by a group at Center for Digital Experience at Pratt Institute. Team members: Aemilia Techentin, Isik Erturk, Jichen Zhu and Lalita Shripad Chavan. The team completed the whole process of cooperation from initial research, client meeting, program planning, to data cleaning, data analysis, create dashboard, conduct findings & recommendations, write report and presentation.
My role: Data cleaning, analyze social media data (Facebook), consolidate recommendations accordingly to the analyzed data, create a dashboard on Google Data Studio, make social media post mockups, write a report, craft a presentation deck and present findings and recommendations to the client/stakeholders.
RESEARCH OBJECTIVE & SCOPE
The goal of this study is to understand The Henry Ford Museum’s website traffic and social media engagement based on data collected by Google Analytics and Sprout Social. This one month of research focuses on analyzing engagement for selected categories: artifacts, events and exhibitions on social media platforms and evaluating overall performance of social media channels to drive traffic to the museum’s website in accordance with the organization’s objective. While enhancing the visibility of the Henry Ford Museum’s website, the analysis will contribute to improvement of future content to attract more users on Facebook, Instagram and Twitter.
METHODOLOGY
PROCESS TIMELINE
GETTING THE DATA
All data in this study was provided by The Henry Ford. During our first meeting we were able to define the scope of the study, which would be looking both at website traffic data and social media interactions on the networks themselves. Given this, The Henry Ford provided us with both 1) Access to the TheHenryFord Combined Site Google Analytics account and 2) all social media analytics data (on all platforms) from Sprout Social collected during the period of January 1 - October 31, 2022. Sprout Social analytics data was accessible to our team via an exported excel sheet file. While it is possible to view data from any time period in Google Analytics, our Sprout Social data limited us to a more specific time period in 2022. For this reason, the period analyzed in this study is Jan 01,2022 - Oct 31, 2022. All findings and recommendations are based on data collected during this time period only.
Update emails and calls with The Henry Ford provided further insight into where to find relevant traffic data on Google Analytics. From these sessions with our client, we were able to find data on the GA specific to all traffic coming to the site, traffic only from social media sources, social media network referral links and landing pages, and campaign names and performance.
Data from Sprout social gave us: Network names, reach, impression, engagement, link names and click rate, post amounts (record count metric), and type of post (post/video/Story, etc.). It allowed us to find other relevant data points for analysis, specifically: post categories (Artifacts/Events/Exhibits), category post amount by network, category post amount overall, category post type, post frequency, and follower count, and traffic average value per post.
After further communication with the client, we determined the data range of social media networks to be limited to Facebook, Instagram and Twitter.
PROCEDURE
Part 1: Social Media
Clean up the social media data exported from Sprout Social on Google Sheets.
-
Aggregate all data and analyze it for a comprehensive overview of social media performance.
1.
Define key metrics for social media performance that would form the basis of all future analysis: Network, Reach, Impressions, Engagements (and Engagement Rate per Impression), Post Link Clicks, Followers, Post (Amount).
2.
Based on client requirements and our initial overall performance findings from Steps 1 and 2, further screen the social media data, and classify it by artifacts, events and exhibits.
-
Label the posts with their corresponding names in the spreadsheet (i.e. event/exhibit/artifact titles were found on the website and matched to the post record on our Google sheet).
3.
Add Event/exhibit/artifact (“category”) post amount by network, overall post amount, and type of post per category to the social media performance key metric list.
4.
Use Google Looker Studio to design a social media overview dashboard in order to further reflect, display and facilitate our current understanding of the data from Sprout Social.
5.
Analyze widgets that show the data according to our key metrics.
6.
Summarize, classify and filter core social media findings and give corresponding recommendations.
7.
For the three social media platforms, create mockups of different kinds of posts to reflect our recommendations.
8.
Part 2: Traffic to the Website from Social Media
Prepare for a Traffic dashboard that focuses on 1) quantifying traffic to the website from social media and 2) qualifying traffic that traffic.
1.
Use Google Looker Studio to create widgets that show both the overall traffic to the site and social media channel traffic’s impact within that.
2.
Define key data points for Google Analytics:
-
All traffic to the site, traffic only from social media sources (overall and by network), social media network referral links and landing pages, and campaign names and performance.
3.
Define key metrics that can be found on Google Analytics:
-
Sessions, Traffic per Post, Avg. Session Duration, Bounce Rate, Goal Conversion Rate, Ecommerce Conversion Rate, Revenue.
4.
Define key metrics from the social media overview:
-
Number of posts (by network).
5.
Blend data from GA with Sprout Social data to create more dashboard widgets that would show a more holistic picture of the value of traffic that each social network brings to the site.
6.
Create widgets on the dashboard that allow THF to monitor traffic to the artifact pages (page path level 4 under “digital collection” page path level 2) as these are strategically relevant to the organization's goals.
7.
Based on this data, write findings and corresponding recommendations. Reference any additional data points from the social media analysis as needed.
-
Use Figma, screenshots and screen recordings to create all mockups.
8.
FINDINGS & RECOMMENDATIONS
OVERALL
Overall, The Henry Ford made 4,089 posts total across Facebook, Instagram, Twitter, and YouTube. Facebook, Twitter, and Instagram were posted to about the same amount throughout the 10 month period (1,155 - 1,480 posts), which is about 5x more than the organization’s YouTube posting amount (THF made 222 posts during this time period).
The total reach of THF across all four platforms is about 579,513 followers, but Facebook followers account for 80% of that number. Facebook has the highest reach and engagement rate per impression and Twitter has half of that. All other social networks, Including Instagram and YouTube, have less than a 1% engagement rate per post impression. Most social media marketing experts agree that a good engagement rate is between 1% to 5% (Sehl & Tien, 2022).
Facebook sends almost 90% of the total social media traffic to the site. Traffic to the site is affected by reach/impressions and engagement rate, but also the limitations of the platform. For example, Facebook allows for unlimited direct URLs to be placed in each post. Instagram does not allow any web links in their posts.
WEBSITE TRAFFIC - GOOGLE ANALYTICS
Finding 1: The quality of user sessions of Instagram channel traffic to the website is high: very high conversion rates, good session duration and pages per session and a low bounce rate.
Based on Google Analytics acquisition reports, Instagram sends more quality sessions to the website than Facebook and Twitter.
Out of all the major social media traffic channels, Instagram traffic has a lower average bounce rate than all users to the site, and the lowest bounce rate of all social media channels (42.51%). It also has the highest goal conversion rate out of any major social media network (62.35%), including Facebook (which only has a 23.95% conversion rate). THF website users from Instagram visit more pages and spend more time than average on each page compared to the top five social media traffic channels. Instagram also sees the second largest number of total transactions on the site despite the very small number of sessions, and accordingly has the highest ecommerce rate out of any social media channel. At 50,600 followers, its reach is nowhere near Facebook’s 465,913 followers, but it is higher than on all other major networks.
Recommendation 1: Invest in a strategy to increase followers on Instagram. Start with a strategy via reel videos or ads (targeting similar audiences to current followers).
A strategy to increase followers on Instagram may mirror strategy on Facebook. Campaigns similar to ri and sg for Facebook would help increase object awareness and ecommerce rates. Launch campaign tracking, and make sure that links are provided as much as possible (more in Recommendation 2). Instagram provides the unique opportunity to engage users with reel videos (Sehl & Cohen, 2022), which are currently prioritized by Meta’s AI and distributed to a wider general audience, outside of subscribers. Create Reels that feature people, movement and audio. Use stickers for text to help tell a story, like THF’s most viewed reel featured in.
Finding 2: Instagram currently sends the least amount of traffic to the website, on average per post.
Between Jan 1 - Oct 31, 2022 THF, the platforms that account for the most social media posts are Twitter, Instagram, and Facebook, with 1310, 1232, and 1155 posts, respectively. The only other major social media network studies, YouTube, lags behind with 222 posts. When looking at traffic to the website, Facebook alone accounts for 88.39% of social media traffic. Instagram sends the least amount of traffic to the website per post: on average there is only one website session per Instagram post, compared to 100 sessions per Facebook post.
Facebook has a much higher reach than Instagram, as noted in Finding 1. Unlike Instagram, Facebook allows posts to contain links that its users can click on and navigate directly to the site. Indeed, most referral links from all social media channels are from Facebook.
Instagram users can only get links directly to the site from the THF Shop , within 24-hr duration Instagram Stories, or through four LinkTree links. While traffic from LinkTree is not included in this study, it is relevant to note that Linktree does not allow users to click directly on content that is in THF posts, as other competitor link-in-bio tools do. Additionally, the links currently listed in LinkTree do not reflect all categories of Instagram posts that users may be responding to. For example, a user that would like to visit the site to find more information about an event will not find a direct link to the site in LinkTree.
Recommendation 2: Give Instagram followers access to more direct links on the website. Continue posting links to the site on Instagram Stories, in the Shop, and consider using a tool that allows users to find links related directly to posts that they like, such as Linkin.bio.
Keep the Shop on the THF account up-to-date on Instagram by ensuring that there are no expired links under pages or products that have become out of stock. If having users click on content-specific content is a goal of THF, consider switching link-in-bio tools to Linkin.bio. If a user spots an artifact or exhibition that interests them, or a specific event that they like to sign-up for, other link-in-bio tools like Later’s Linkin.Bio allows users to follow the link, and click on a website link that looks like the post they are interested in. Before switching bio linking tools, we recommend analyzing the traffic and conversions that result in LinkTree user clicks during the same time period as this study.
SOCIAL MEDIA
Finding 3 (Facebook): Facebook Engagement could be improved with more active content, increased amounts of photos and videos, and increasing interactivity for users.
More than half (55.4%) of all Events/Exhibits/Artifacts posts on Facebook were photos. Posts in the form of photos have the highest average likes (265.28) and video posts have the highest average engagement rate (3.98%). But overall the average engagement rate of these posts is very low.
Facebook Engagement could be improved with more active content, increased amounts of photos and videos, and increasing interactivity for users. Currently, the content of posts related to exhibitions and events is often repeated and the number of related pictures is small. Photos tend to be missing a focal point that stars as the subject of interest.
Many posts about displaying a series of artifacts only put a picture of one of the artifacts.
Posts that mention both the artifact and the corresponding exhibition/event are often missing or only contain a link to either the artifact or the exhibition/event.
Recommendation 3 (Facebook): Post enriched & interesting content and high- quality photos & videos to attract audiences to interact and gain conversion.
The selection of photos and videos content should be rich and interesting, which can arouse the interest of the audience, so as to click the link to visit The Henry Ford website to browse, purchase tickets, etc. For example, event posts can include more photos of happy moments when people participate in the event and photos of the highlights of the event, etc. Asking users to participate in the engagement(like, reaction, comment, click links) and do some lucky draws or free tickets as rewards, which will arouse their interest and bring some traffic and conversion.
In order to attract more young users, some special content such as AR/VR immersive experience, short video challenge, gift giving, etc. can be posted.
Artifacts posts can post more pictures of relevant details and include links to detailed introduction pages and links to purchase tickets for related events/exhibitions.
Finding 4 (Instagram): Facebook Engagement could be improved with more active content, increased amounts of photos and videos, and increasing interactivity for users.
Photos compose the greatest amount of the all posted content on Instagram with 72%. Carousels are the second most popular content type, following up with 21%. Comparing these two categories, one expects their impacts on likes and impressions to be quite different as well. Nevertheless, in all three bar graphs, they averaged very similarly, in most cases carousels even beating photos. This leads us to the conclusion that although Henry Ford Museum did not post many carousels, doing that could have supported the overall engagement rates of the network.
We also took a detailed look at the videos of Instagram. 80% of the total video content was posted as stories whereas reels’ percentage was as small as 1%. When the average likes and engagement rate is taken into account, reels happens to be the content type that attracts the most. Likes that reels receive is almost the double of the amount of likes received from posts.
Looking at the posts with a qualitative perspective, Henry Ford Museum chose to demonstrate the visual artifacts solely by themselves either in the natural atmosphere of the exhibition or with emphasis on the object with a clear background. Lastly, were generally composed of 3 paragraphs, first two paragraphs devoted to the explanation of the item while the last one informing us about its classification. Moreover, not every post included hashtags. Only across exhibition posts the hashtag usage was prominent.
Recommendation 4 (Instagram): Posting more carousels and Reels with visuals showing interaction and shorter captions that encourage engagement.
Depending on our findings, the Henry Ford Museum don’t post many carousels and reels about Artifacts/Events/Exhibitions but there is a great opportunity for them to start doing so, as they are very effective on engagement rates.
Other than implementing these main outcomes into design, there are basic Instagram recommendations that can be taken into account. Since Instagram is a platform that prioritizes visuals, the goal should always be posting interesting pictures with consistent dimensions to follow the rules of design guidelines and keeping short captions because longer the textual content gets, higher the chance of more users to miss the full message.
For the visuals, including photographs from the exhibitions that show how the viewers are interacting with artifacts, rather than simply presenting the artifact, can encourage social media followers to go and see the exhibitions in person. Other than that, digital interaction with the followers should be aimed as comments improve engagement. This can be done by asking an inviting question about the artifact at the very end of the caption. By this way, followers can participate more rather than just being mere viewers.
The current home page of Instagram consists of posts that differentiate in their visuals’ sizing. While creating these stories, we want to remind that following a system for design decisions should always be the key for a more organized and coherent look.
Finding 5 (Instagram): Since most of the stories on Instagram are not saved, their links in the data do not work.
Instagram stories are built to disappear after 24 hours. Therefore, although there were many stories data on the Social Sprout sheet posted as stories, we could not really study them as they expired and their links did not work anymore. We also tried the saved stories but most of them were not up to date.
Recommendation 5 (Instagram): Create saved stories categories for selected categories and try keeping them updated as much as possible.
To be able to reach to the stories of the specific contents that we want to keep track of, we are suggesting to manage saved stories categories. We realized that there is already an existing saved stories categories for “Collections” and “Exhibitions”. Unfortunately, they were not updated for around 44 weeks. Keeping that one up to date and creating new categories for “Events” and “Exhibits” to collect the published stories permanently would both eliminate the risk of losing their data, create an archive for the interested users to be able to discover or learn more and guarantee their visibility.
Finding 6 (Twitter): The average engagements and avg impressions are the second highest for video content but the number of video posts account to only 0.2% of the total posts.
The highest average engagements of 15 (where, engagements = number of reactions + number of comments + number of shares + number of post clicks(all)), average impressions of 779 and average post link clicks of 532 come from photo posts. But they also account for a huge chunk of the total posts which is 71.8%.
On the other hand, video posts account for only 0.2% of the total posts (if we look at the number, only 1 video post has been posted for the selected
content categories and the selected time frame). But the single video post brings in the second highest average engagements of 9, average impressions of 536.
Recommendation 6 (Twitter): Include video content when posting about an artifact/exhibit/event to improve engagements and impressions which leads to a chance of improving traffic to the website.
Including a video while posting about an artifact/exhibit/event along with the related website link on Twitter will help improve user engagement and impressions on the posts. This will also eventually lead to an increase in traffic to the website.
Finding 7 (Twitter): The exhibits category posts bring in more traffic to the Henry Ford website as compared to the artifacts category posts.
The number of artifact posts account for almost 48% of the total posts and the total post link clicks is 5,970, which is less than half the amount if we compare it with the exhibit category posts. The number of exhibit posts account for 31% of the total posts and the total post link clicks is 14,455.
Recommendation 7 (Twitter): Increase the number of exhibits category posts and include links to the related artifacts.
Since the exhibit category posts are performing better than the artifacts category posts, increasing the number of exhibit posts that contain links to the artifacts that are related to the exhibit will help direct traffic to those artifact pages on the website.
Also, including compelling and gripping content along with adding up to 2 hashtags for the posts will help improve engagements and reach for the posts.
DASHBOARD
CONCLUSION
OVERALL
Our goal of this study was to match up web traffic data for The Henry Ford, with social media network engagement and overall value. Another goal of ours was to drill down on the content that is posted on social media that is tied specifically to museum artifacts, their events, and exhibitions.
Currently most of the overall traffic to The Henry Ford website is not coming from social media channels. Most of the traffic that does make it to the site from social networks, is coming exclusively from Facebook. However, conversion rate data from Google Analytics tells us that it is worthwhile for THF to invest in Instagram and make improvements to their posting strategies on three major social media networks: Facebook, Instagram, and Twitter.
Across all of its social media networks, most of THF’s reach comes from Facebook, despite sending more posts to Instagram and Twitter. Ways to increase reach on Instagram, such as creating effective Instagram Reels, may be a good first step in strategizing for future growth on a platform beside Facebook. Instagram audiences have the highest conversion rates on the website, and it is recommended that THF work more direct links into the content they post to Instagram, via updated Shop items, more links in Stories, and perhaps a link-in-bio tool that matches posts to website URLs. Currently, engagement on the app itself is still lacking so, in addition to gaining viewers and engagements on Reels, THF should focus on high quality images and videos on the site.
We think it is valuable to stay ahead of the curve on Facebook, too, although it is the organization’s number one leader in driving traffic to the site and has the largest reach. Engagement still has room for improvement; more active content, increased amounts of photos and videos, and increasing interactivity for users is found to lead to higher engagement on the platform.
A NOTE ON LIMITATIONS & NEXT STEPS
Our study faced two main constraints: We were not able to use “Like” data from all networks and not all Instagram website traffic was able to be analyzed.
Sprout social does not have accurate “Like” data for Facebook; manually clicking into the Facebook posts revealed that the “Likes” were sometimes more than twice as much as reported on the sheet. Additionally, because visitors from Instagram will go through an intermediary (i.e., Linktree), they will not be recorded as coming from “Social – Instagram” but instead as “Social – Other” in Google analytics (Mores 2022).
For next steps:
RESULTS: THE CLENT WAS IMPRESSED!
