LinkedIn Analytics Dashboard

Build me a working, private LinkedIn Analytics Dashboard from my own LinkedIn analytics export.

Create the actual interactive dashboard with my results, charts and insights already inside it. I want a finished dashboard I can open and use, not just a written report, spreadsheet, screenshot or code snippets.

I will attach my personal-profile LinkedIn analytics Excel export (.xlsx), ideally covering the last 90 days. If it is missing, ask me to upload it before building. If I am uploading multiple files, wait until I say all files are uploaded.

Read every worksheet and identify the available dates, account-level metrics, post rankings, follower data and demographic snapshots. Sheet names may include DISCOVERY, ENGAGEMENT, TOP POSTS, FOLLOWERS, AUDIENCE DEMOGRAPHICS and CONTENT DEMOGRAPHICS; inspect the actual structure rather than assuming those names or columns always exist.

Use only my supplied figures. Do not insert demonstration data, reuse another person's results or invent missing metrics. Keep the original workbook unchanged. If a section cannot be supported, show a short explanation of what is missing and build the supported sections.

DESIGN AND LAYOUT

Create an elegant, mobile-friendly report dashboard using only neutral monochrome colours throughout:
• Off-white #FAFAF8 — page background.
• White #FFFFFF — cards and panels, and text on charcoal buttons.
• Charcoal #202020 — headings, main text, buttons, links and primary chart series.
• Mid-grey #6B6B6B — secondary text and secondary chart series.
• Light grey #E3E3E0 — borders, chart tracks and subtle callout backgrounds.

Keep normal text charcoal or mid-grey on white or off-white backgrounds. Use charcoal text on light-grey callouts. Use accessible contrast, clear focus indicators, labelled controls and comfortable tap targets. Chart series must also have written labels and values; do not rely on colour alone.

Use a centred desktop layout around 1,160px wide, clean sans-serif typography, generous spacing, thin borders and softly rounded panels. Main headings should be around 34–53px depending on screen size. Keep the introduction compact enough that the summary figures appear quickly. On phones, stack the main sections and use a two-by-two summary-card grid. Keep wide data tables scrollable inside their own panels.

At the top, show my name if it is clearly available in the upload; otherwise use “Your LinkedIn review”. Include the actual export period, a private-report label and a Print report button.

Add navigation links to Overview, Momentum, Posts, Audience, Your next month, and Sources & method. Use those sections in that order.

1. OVERVIEW

Write an evidence-led headline about the strongest finding in MY data, a short plain-English explanation, and one useful working hypothesis to test. Do not copy a conclusion from someone else's dashboard.

Show four summary cards where the data supports them:
• Impressions for the exported period.
• Members reached for the exported period.
• New followers reported during that period.
• Total followers at the available snapshot date.

Explain each label briefly. New follower additions are not automatically net follower growth. Total followers is a snapshot, not a period total. Do not invent a growth percentage without a valid opening value and a compatible closing value. State the scope of the cards; chart filters should not silently change them.

2. MOMENTUM

Compare the latest complete 28-day window with the immediately preceding complete 28-day window, when both are present. Exclude dates known to be partial and label the actual boundaries. If there are not enough days, explain this rather than manufacturing a comparison.

Show impressions and new followers with both totals, percentage changes and paired horizontal comparison bars. Handle a zero baseline honestly rather than showing an infinite percentage. Explain what the change might mean without claiming that particular posts caused follower growth.

Add an interactive activity chart with:
• Period: full export, latest complete 28 days, previous complete 28 days where available.
• Metric: impressions or new followers where available.
• View: daily or weekly.
• A selected-period total and date range.
• A slider to inspect individual days or weeks, with a written date and value.
• An expandable daily-data table, including signed engagement changes if supplied.

Use Monday-based weeks, label partial first and last weeks, and preserve gaps instead of treating missing dates as zero. All controls must update the correct series, labels, totals and inspection value.

3. POSTS

Create up to three evidence-led case-study cards: strongest reach, a useful engagement-per-impression example, and a smaller-audience example worth testing. Choose genuinely different examples where possible; fewer cards are fine if the data is sparse.

Each card should show the available post title or readable URL opening, its actual figures, a short interpretation and a working link to the post. Label URL-derived titles as such. If full post text or visuals are not provided, do not pretend you have read or seen them or identify formats with certainty.

Add a searchable, sortable post table with:
• Post title or URL opening and clickable link.
• Publication date, only where supported by reliable metadata; otherwise “Not available”.
• Impressions.
• Engagements.
• Engagements divided by impressions, shown as a clearly labelled descriptive ratio when valid.

Search the available text. Allow sorting by impressions, engagements, ratio and publication date where supported. Offer “Published during the export period” and “All listed posts” only when dates can be established. Show the first 10 results with a Show all / Show first 10 control and matching-result counts. Provide a useful empty state.

Merge separate top-post rankings by exact post URL. Preserve missing metrics as “Not listed”, not zero. A top-post ranking is not necessarily the full publication history. Keep posts published before the reporting period identifiable.

The ratio is not the percentage of unique people who engaged. Do not calculate it with zero or missing impressions. Flag examples below 500 impressions so small-sample ratios are not oversold. Explain that posts can differ in age and metric coverage.

4. AUDIENCE

Show up to three specific, useful findings supported by the demographic data, with cautious ideas for relevant content.

Create paired horizontal demographic bars for followers and content demographics where both exist. Keep the populations separately labelled. Do not rename content demographics “buyers” or “engaged followers” without evidence.

Add a dimension selector for the available fields, such as industry, job title, location, seniority, company size and company. Include a legend, written percentages, the first eight categories and a Show all / Show first eight control. If only one population is supplied, show that one honestly.

Retain rounded values and “<1%” as bounds. Missing categories are “Not listed”, not zero. Do not add overlapping categories or assume a partial ranking totals 100%. Keep snapshot dates visible where available.

5. YOUR NEXT MONTH

Translate the findings into three focused content priorities, each tied to evidence and a practical action. Distinguish observations from hypotheses. Avoid generic advice such as “post consistently” without a useful explanation.

Add four working tabs: Week 1, Week 2, Week 3 and Week 4. Each week should have a clear aim and three tailored post ideas, each with a proposed angle or opening and a short explanation of what to test. These are suggested future ideas, not claims about posts I have already published.

Add a small measurement plan covering comparable-age post impressions, engagement-per-impression where valid, follower additions and any manually tracked enquiries. Explain which outcomes the export cannot establish. Do not promise growth or attribute leads to a post without evidence.

6. SOURCES & METHOD

List the actual uploaded filenames, sheets read, date coverage and relevant limitations. Reconcile daily totals against period totals where the measures and dates match; report unexplained differences rather than silently changing data.

Preserve negative daily engagement values and describe them as signed changes. Do not turn them into gross engagement counts, reset them to zero or mix daily account-level totals with post-ranking totals to create a misleading headline engagement rate.

Do not declare a best posting time without publication-time data. If discussing weekday visibility, explain that viewing dates are not publishing dates and check whether one unusual day dominates the result. Do not claim causal effects from descriptive patterns.

PRIVACY, PRINTING AND DELIVERY

Treat this as an export snapshot, not a live LinkedIn connection. Do not add a refresh or account-sync button unless that connection is actually implemented and tested. For later updates, I can upload a fresh export in this conversation and ask you to rebuild the dashboard using it.

Keep my personal analytics private. Use private hosting when supported; otherwise deliver a working local or private artifact supported by this environment. Do not publish my analytics publicly. If the environment cannot create a working dashboard, explain the limitation instead of claiming it has done so.

Make Print report produce a readable report with the full post list, all four weekly plans and the relevant methodology visible. Hide controls that do not belong in print. Support keyboard operation and reduced-motion preferences.

Before delivery, check the calculations, missing values, period boundaries, ranking joins, filters, search, sorting, demographic selector, week tabs and print behaviour. Check mobile layout if browser tools are available. State which checks were performed and any limits.

Give me the finished dashboard link or file and a short explanation of how to use it. Start by checking my uploaded export.