Total Jobs Scraped

Salary Landscape

Salary Distribution (USD and PHP across different salary types)

df_valid_salary = df.dropna(subset=["average_salary", "currency", "salary_type"]).copy()
df_valid_salary = df_valid_salary[df_valid_salary["salary_type"] != "unknown"]
df_valid_salary = df_valid_salary[df_valid_salary["salary_type"] != "annually"]
df_valid_salary = df_valid_salary[df_valid_salary["salary_type"] != "daily"]

currencies = df_valid_salary["currency"].unique()
salary_types = df_valid_salary["salary_type"].unique()

n_bins = int(2 * len(df) ** (1 / 3))

for curr in currencies:
    for sal_type in salary_types:
        subset = df_valid_salary[
            (df_valid_salary["currency"] == curr)
            & (df_valid_salary["salary_type"] == sal_type)
        ]

        if not subset.empty:
            color = "#636EFA" if curr == "USD" else "#FFA15A"

            fig = px.histogram(
                subset,
                x="average_salary",
                title=f"Salary Distribution: {curr} - {sal_type.capitalize()}",
                color_discrete_sequence=[color],
                nbins=n_bins,
                height=500,
            )

            fig.update_layout(
                xaxis_title=f"Average Salary ({curr})",
                yaxis_title="Number of Listings",
                template="plotly_white",
            )
            fig.show()

Analysis

Disclaimer: all the analysis in this notebook was generated by gemini 3.1 pro by feeding the visualizations to it.

PHP Monthly: Heavily right-skewed with spikes at clean numbers — 20k, 30k, 40k, 50k. Most listings fall between 20k–60k PHP, with a long tail above 100k.

USD Monthly: Strong clustering at $500 and $1,000 , with other spikes at $400, $800, and $1,500.

Pay Range By work type (usd and php across different salary types)

IQR box plots showing median, Q1–Q3, and outlier range

for curr in currencies:
    for sal_type in salary_types:
        subset = df_valid_salary[
            (df_valid_salary["currency"] == curr)
            & (df_valid_salary["salary_type"] == sal_type)
        ]

        if not subset.empty:
            fig = px.box(
                subset,
                x="work_type",
                y="average_salary",
                color="work_type",
                title=f"Pay Range by Work Type: {curr} - {sal_type.capitalize()}",
                height=600,
                points="outliers",  
            )

            fig.update_layout(
                template="plotly_white",
                xaxis_title="Work Type",
                yaxis_title=f"Average Salary ({curr})",
                legend_title="Work Type",
            )
            fig.show()

Analysis

PHP Monthly: Full-time leads with the highest median (~45k PHP) and widest range (30k–60k). Part-time and Any have lower medians (~35k and ~25k) but both show high-earning outliers up to 100k. Gig work sits around 20k; Freelance listings are sparse, clustering near 15k.

USD Monthly: Full-time leads with a median of ~$800 and a range of $600–$1,200. Other categories (Part-time, Any, Gig) have lower medians ($450–$700), but all feature dense clusters of outliers reaching $2,000 — matching the absolute peak of full-time earners.

Average Pay By Category (Top 10)

cat_df = df.dropna(subset=["category"])
cat_df = cat_df[cat_df["salary_type"] == "monthly"]

cat_group = (
    cat_df.groupby("category")
    .agg(avg_usd=("usd_salary", "mean"))
    .reset_index()
)

top10_cat = cat_group.nlargest(10, "avg_usd").sort_values("avg_usd", ascending=True)

fig_usd = go.Figure(
    go.Bar(
        y=top10_cat["category"],
        x=top10_cat["avg_usd"],
        name="Avg USD Salary",
        orientation="h",
        marker_color="steelblue",
        text=top10_cat["avg_usd"].map(lambda x: f"${x:,.0f}"),
        textposition="outside",
    )
)
fig_usd.update_layout(
    title="Average Monthly Pay by Category (Top 10) — USD",
    xaxis_title="Average Monthly Salary (USD)",
    yaxis_title="Category",
    height=500,
    margin=dict(l=200, r=150),
)
fig_usd.show()

Average Pay by SubCategory (Top 10) (both usd and php)

sub_df = df.dropna(subset=["subcategory"])
sub_df = sub_df[
    (sub_df["salary_type"] == "monthly") & (sub_df["work_type"] == "Full Time")
]

sub_group = (
    sub_df.groupby("subcategory")
    .agg(avg_usd=("usd_salary", "mean"))
    .reset_index()
)

top10_sub = sub_group.nlargest(10, "avg_usd").sort_values("avg_usd", ascending=True)

fig_usd = go.Figure(
    go.Bar(
        y=top10_sub["subcategory"],
        x=top10_sub["avg_usd"],
        name="Avg USD Salary",
        orientation="h",
        marker_color="steelblue",
        text=top10_sub["avg_usd"].map(lambda x: f"${x:,.0f}"),
        textposition="outside",
    )
)
fig_usd.update_layout(
    title="Average Monthly Pay by Subcategory (Top 10) — USD",
    xaxis_title="Average Monthly Salary (USD)",
    yaxis_title="Subcategory",
    height=550,
    margin=dict(l=250, r=150),
)
fig_usd.show()

Market Demands

def parse_skills(series):
    return series.dropna().str.split(";").explode().str.strip().str.lower()

Top 10 Required Skills

Skills listed most frequently across all job postings

skill_counts = parse_skills(df["skills"]).value_counts().head(10).reset_index()
skill_counts.columns = ["Skill", "Count"]

fig_top_skills = px.bar(
    skill_counts,
    x="Count",
    y="Skill",
    orientation="h",
    title="Top 10 Most Required Skills",
    text="Count",
    color="Count",
    color_continuous_scale="Blues",
)

fig_top_skills.update_layout(
    yaxis=dict(autorange="reversed"),
    xaxis_title="Number of Job Listings",
    yaxis_title="Skill",
    coloraxis_showscale=False,
    template="plotly_white",
)
fig_top_skills.update_traces(textposition="outside")
fig_top_skills.show()

Skills by Work Type

Which skills dominate each work arrangement

work_types = df["work_type"].dropna().unique()

for wt in work_types:
    subset = df[df["work_type"] == wt]
    skill_wt_counts = (
        parse_skills(subset["skills"]).value_counts().head(10).reset_index()
    )
    skill_wt_counts.columns = ["Skill", "Count"]

    if skill_wt_counts.empty:
        continue

    fig_wt = px.bar(
        skill_wt_counts,
        x="Count",
        y="Skill",
        orientation="h",
        title=f"Top 10 Skills — {wt}",
        text="Count",
        color="Count",
        color_continuous_scale="Oranges",
    )

    fig_wt.update_layout(
        yaxis=dict(autorange="reversed"),
        xaxis_title="Number of Job Listings",
        yaxis_title="Skill",
        coloraxis_showscale=False,
        template="plotly_white",
    )
    fig_wt.update_traces(textposition="outside")
    fig_wt.show()

Skill Pairing Heatmap

How often top skills appear together in the same listing

TOP_N = 15  

top_skills = parse_skills(df["skills"]).value_counts().head(TOP_N).index.tolist()

pair_counts = Counter()

for skill_list in df["skills"].dropna():
    skills_in_row = list(
        {
            s.strip().lower()
            for s in skill_list.split(";")
            if s.strip().lower() in top_skills
        }
    )
    for pair in combinations(sorted(skills_in_row), 2):
        pair_counts[pair] += 1

matrix = pd.DataFrame(0, index=top_skills, columns=top_skills)

for (s1, s2), count in pair_counts.items():
    matrix.loc[s1, s2] = count
    matrix.loc[s2, s1] = count

fig_heatmap = go.Figure(
    data=go.Heatmap(
        z=matrix.values,
        x=matrix.columns.tolist(),
        y=matrix.index.tolist(),
        colorscale="Purples",
        text=matrix.values,
        texttemplate="%{text}",
        hoverongaps=False,
        hovertemplate="<b>%{y}</b> + <b>%{x}</b><br>Co-occurrences: %{z}<extra></extra>",
    )
)

fig_heatmap.update_layout(
    title=f"Skill Pairing Heatmap (Top {TOP_N} Skills)",
    xaxis_title="Skill",
    yaxis_title="Skill",
    template="plotly_white",
    height=650,
    xaxis=dict(tickangle=-35),
)

fig_heatmap.show()

Top 50 Required Skills

Skills listed most frequently across all job postings (regardless of work type)

skill_counts_50 = parse_skills(df["skills"]).value_counts().head(50).reset_index()
skill_counts_50.columns = ["Skill", "Count"]

fig_top50 = px.bar(
    skill_counts_50,
    x="Count",
    y="Skill",
    orientation="h",
    title="Top 50 Most Required Skills",
    text="Count",
    color="Count",
    color_continuous_scale="Blues",
    height=1400,
)

fig_top50.update_layout(
    yaxis=dict(autorange="reversed"),
    xaxis_title="Number of Job Listings",
    yaxis_title="Skill",
    coloraxis_showscale=False,
    template="plotly_white",
)
fig_top50.update_traces(textposition="outside")
fig_top50.show()

Skill Pairing Heatmap (Top 25 Skills)

How often top 25 skills appear together in the same listing

TOP_N_25 = 25

top_skills_25 = parse_skills(df["skills"]).value_counts().head(TOP_N_25).index.tolist()

pair_counts_25 = Counter()

for skill_list in df["skills"].dropna():
    skills_in_row = list(
        {
            s.strip().lower()
            for s in skill_list.split(";")
            if s.strip().lower() in top_skills_25
        }
    )
    for pair in combinations(sorted(skills_in_row), 2):
        pair_counts_25[pair] += 1

matrix_25 = pd.DataFrame(0, index=top_skills_25, columns=top_skills_25)

for (s1, s2), count in pair_counts_25.items():
    matrix_25.loc[s1, s2] = count
    matrix_25.loc[s2, s1] = count

fig_heatmap_25 = go.Figure(
    data=go.Heatmap(
        z=matrix_25.values,
        x=matrix_25.columns.tolist(),
        y=matrix_25.index.tolist(),
        colorscale="Purples",
        text=matrix_25.values,
        texttemplate="%{text}",
        hoverongaps=False,
        hovertemplate="<b>%{y}</b> + <b>%{x}</b><br>Co-occurrences: %{z}<extra></extra>",
    )
)

fig_heatmap_25.update_layout(
    title="Skill Pairing Heatmap (Top 25 Skills)",
    xaxis_title="Skill",
    yaxis_title="Skill",
    template="plotly_white",
    height=850,
    width=950,
    xaxis=dict(tickangle=-35),
)

fig_heatmap_25.show()

Work Types & Trends

Break Down of how many jobs per work type

work_type_counts = df["work_type"].value_counts().reset_index()
work_type_counts.columns = ["Work Type", "Job Count"]

fig1 = px.pie(
    work_type_counts,
    names="Work Type",
    values="Job Count",
    title="Number of Jobs per Work Type",
    color_discrete_sequence=px.colors.qualitative.Pastel,
)

fig1.update_traces(textposition="inside", textinfo="percent+label")
fig1.update_layout(template="plotly_white")
fig1.show()

Over Time Posting Volume by date per month

df["date_updated_parsed"] = pd.to_datetime(df["date_updated"], errors="coerce")
df["year_month"] = df["date_updated_parsed"].dt.to_period("M")

monthly_volume = (
    df.dropna(subset=["date_updated_parsed"])
    .groupby("year_month")
    .size()
    .reset_index(name="job_count")
    .sort_values("year_month")
)
monthly_volume["year_month_dt"] = monthly_volume["year_month"].dt.to_timestamp()

fig = go.Figure()

fig.add_trace(
    go.Scatter(
        x=monthly_volume["year_month_dt"],
        y=monthly_volume["job_count"],
        mode="lines+markers+text",
        line=dict(color="#4C72B0", width=2.5),
        marker=dict(size=8, color="white", line=dict(color="#4C72B0", width=2.5)),
        text=monthly_volume["job_count"],
        textposition="top center",
        textfont=dict(size=11),
        fill="tozeroy",
        fillcolor="rgba(76, 114, 176, 0.12)",
        hovertemplate="<b>%{x|%B %Y}</b><br>Job Postings: %{y}<extra></extra>",
    )
)

fig.update_layout(
    title=dict(
        text="Job Posting Volume Over Time (Monthly)",
        font=dict(size=16),
        x=0.5,
        xanchor="center",
    ),
    xaxis=dict(
        title="Month", tickformat="%b %Y", tickangle=-45, showgrid=False, dtick="M1"
    ),
    yaxis=dict(title="Number of Job Postings", showgrid=True, gridcolor="#eeeeee"),
    plot_bgcolor="white",
    hovermode="x unified",
    margin=dict(t=60, b=80, l=60, r=40),
)

fig.show()

Weekly Hours Distribution (hours pattern)

hours_clean = pd.to_numeric(df["hours_per_week"], errors="coerce").dropna()
hours_clean = hours_clean[hours_clean > 0]

bins = [0, 10, 20, 30, 40, 50, 60, 80, float("inf")]
labels = ["1-10", "11-20", "21-30", "31-40", "41-50", "51-60", "61-80", "80+"]
hours_binned = pd.cut(hours_clean, bins=bins, labels=labels)
hours_dist = (
    hours_binned.value_counts().reindex(labels).fillna(0).astype(int).reset_index()
)
hours_dist.columns = ["range", "count"]

colors = ["#E05A2B" if r == "31-40" else "#4C72B0" for r in hours_dist["range"]]

fig2 = go.Figure()

fig2.add_trace(
    go.Bar(
        x=hours_dist["range"],
        y=hours_dist["count"],
        marker_color=colors,
        marker_line_color="white",
        marker_line_width=1.5,
        text=hours_dist["count"],
        textposition="outside",
        hovertemplate="<b>%{x} hrs/week</b><br>Job Postings: %{y}<extra></extra>",
    )
)

fig2.update_layout(
    title=dict(
        text="Weekly Hours Distribution<br><sup>TBD & missing values excluded</sup>",
        font=dict(size=16),
        x=0.5,
        xanchor="center",
    ),
    xaxis=dict(title="Hours per Week", showgrid=False),
    yaxis=dict(title="Number of Job Postings", showgrid=True, gridcolor="#eeeeee"),
    plot_bgcolor="white",
    margin=dict(t=80, b=60, l=60, r=40),
    annotations=[
        dict(
            x=0.99,
            y=0.97,
            xref="paper",
            yref="paper",
            text="31-40 hrs = standard full-time",
            showarrow=False,
            font=dict(size=11, color="#E05A2B"),
            align="right",
        ),
        dict(
            x=0.99,
            y=0.90,
            xref="paper",
            yref="paper",
            text="🔵 Other ranges",
            showarrow=False,
            font=dict(size=11, color="#4C72B0"),
            align="right",
        ),
    ],
)

fig2.show()

print(f"Records with valid hours: {len(hours_clean):,} / {len(df):,} total")

Records with valid hours: 20,147 / 23,908 total

Pay Vs Hours

scatter plot for pay vs hours in USD and in PHP

df["hours_per_week"] = pd.to_numeric(df["hours_per_week"], errors="coerce")
df = df.dropna(subset=["hours_per_week", "average_salary"])
df = df[
    df["hours_per_week"]
    == df["hours_per_week"].apply(lambda x: int(x) if pd.notna(x) else x)
]
df = df[df["salary_type"] == "monthly"]

work_types = df["work_type"].unique()
colors = {
    wt: clr
    for wt, clr in zip(
        work_types, ["#636EFA", "#EF553B", "#00CC96", "#AB63FA", "#FFA15A"]
    )
}

for currency in ["USD", "PHP"]:
    subset = df[df["currency"] == currency]

    fig = go.Figure()

    for wt in work_types:
        wt_data = subset[subset["work_type"] == wt]
        fig.add_trace(
            go.Scatter(
                x=wt_data["hours_per_week"],
                y=wt_data["average_salary"],
                mode="markers",
                name=wt,
                marker=dict(size=9, color=colors.get(wt), opacity=0.75),
                hovertemplate=(
                    f"<b>{wt}</b><br>"
                    "Hours/Week: %{x}<br>"
                    f"Avg Salary ({currency}): %{{y:,.2f}}<br>"
                    "%{text}<extra></extra>"
                ),
                text=wt_data["title"].str.slice(0, 60),
            )
        )

    fig.update_layout(
        title=f"Pay vs Hours — {currency}",
        xaxis_title="Hours per Week",
        yaxis_title=f"Average Salary ({currency})",
        legend_title="Work Type",
        template="plotly_white",
        hovermode="closest",
        legend=dict(itemclick="toggle", itemdoubleclick="toggleothers"),
    )

    fig.show()

Analysis

Market Overview: Volume, Structure & Hours

Job Volume Over Time: Postings peaked in October 2025 (~7,000), dipped to a low of 4,153 in December , then rebounded strongly in January 2026 (6,200+).

Work Type Breakdown: Full-time dominates at 51.4%, followed by part-time (25.2%), flexible/Any (17.5%), and gig (5.8%).

Hours Distribution: The 31–40 hr/week bracket overwhelmingly leads (10,653 postings), with 11–20 hrs as a distant second (3,927). Roles outside these two ranges are uncommon.

Pay vs. Hours: Both USD and PHP scatter plots show dense clusters at 40 hrs (full-time) and 20 hrs (part-time). The wide vertical spread at 40 hours confirms that pay varies more by role and skill level than by hours worked. PHP part-time roles mostly fall under 40,000 PHP; full-time spans 20,000–100,000+ PHP.

Top Skills & Pairings: Video editing leads all skills (2,108 listings), and the entire top five is digital media-focused. Skill pairing heatmaps show three tight clusters: video production (video editing + social media video editing, 531 pairings), digital marketing (Facebook ads + Google ads, 294 pairings), and social media (management + content creation, 285 pairings).

END