diff --git a/action.yml b/action.yml
index 267990ef5..50d5379e1 100644
--- a/action.yml
+++ b/action.yml
@@ -60,6 +60,9 @@ inputs:
   fail-threshold:
     description: 'Threshold which determines if the current workflow fails. Format is the same as alert-threshold input. If this value is not specified, the same value as alert-threshold is used'
     required: false
+  compare-with-best:
+    description: 'Use best result instead of previous result for computing ratio'
+    required: false
   alert-comment-cc-users:
     description: 'Comma separated GitHub user names which start with @ (e.g. "@foo,@bar"). They will be mentioned in commit comment for alert.'
     required: false
diff --git a/src/config.ts b/src/config.ts
index 8e18f2199..1503b3f05 100644
--- a/src/config.ts
+++ b/src/config.ts
@@ -30,6 +30,7 @@ export interface Config {
     alertThreshold: number;
     failOnAlert: boolean;
     failThreshold: number;
+    compareWithBest: boolean;
     alertCommentCcUsers: string[];
     externalDataJsonPath: string | undefined;
     maxItemsInChart: number | null;
@@ -239,6 +240,7 @@ export async function configFromJobInput(): Promise<Config> {
     const commentOnAlert = getBoolInput('comment-on-alert');
     const alertThreshold = getPercentageInput('alert-threshold');
     const failOnAlert = getBoolInput('fail-on-alert');
+    const compareWithBest = getBoolInput('compare-with-best');
     const alertCommentCcUsers = getCommaSeparatedInput('alert-comment-cc-users');
     let externalDataJsonPath: undefined | string = core.getInput('external-data-json-path');
     const maxItemsInChart = getUintInput('max-items-in-chart');
@@ -280,6 +282,7 @@ export async function configFromJobInput(): Promise<Config> {
         commentOnAlert,
         alertThreshold,
         failOnAlert,
+        compareWithBest,
         alertCommentCcUsers,
         externalDataJsonPath,
         maxItemsInChart,
diff --git a/src/write.ts b/src/write.ts
index 856f3ec88..d1bbdd06f 100644
--- a/src/write.ts
+++ b/src/write.ts
@@ -89,12 +89,22 @@ interface Alert {
     ratio: number;
 }
 
-function findAlerts(curSuite: Benchmark, prevSuite: Benchmark, threshold: number): Alert[] {
-    core.debug(`Comparing current:${curSuite.commit.id} and prev:${prevSuite.commit.id} for alert`);
+function findAlerts(
+    curSuite: Benchmark,
+    prevSuite: Benchmark,
+    prevBest: { [key: string]: BenchmarkResult },
+    threshold: number,
+    compareWithBest: boolean,
+): Alert[] {
+    if (compareWithBest) {
+        core.debug(`Comparing current:${curSuite.commit.id} and prev:${prevSuite.commit.id} for alert`);
+    } else {
+        core.debug(`Comparing current:${curSuite.commit.id} and best results for alert`);
+    }
 
     const alerts = [];
     for (const current of curSuite.benches) {
-        const prev = prevSuite.benches.find((b) => b.name === current.name);
+        const prev = compareWithBest ? prevBest[current.name] : prevSuite.benches.find((b) => b.name === current.name);
         if (prev === undefined) {
             core.debug(`Skipped because benchmark '${current.name}' is not found in previous benchmarks`);
             continue;
@@ -141,7 +151,10 @@ function floatStr(n: number) {
     return n.toString();
 }
 
-function strVal(b: BenchmarkResult): string {
+function strVal(b: BenchmarkResult | undefined): string {
+    if (b === undefined) {
+        return '';
+    }
     let s = `\`${b.value}\` ${b.unit}`;
     if (b.range) {
         s += ` (\`${b.range}\`)`;
@@ -157,28 +170,36 @@ function commentFooter(): string {
     return `This comment was automatically generated by [workflow](${actionUrl}) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`;
 }
 
-function buildComment(benchName: string, curSuite: Benchmark, prevSuite: Benchmark): string {
+function buildComment(
+    benchName: string,
+    curSuite: Benchmark,
+    prevSuite: Benchmark,
+    prevBest: { [key: string]: BenchmarkResult },
+    compareWithBest: boolean,
+): string {
+    const ratioStr = compareWithBest ? 'Ratio vs. Best' : 'Ratio';
     const lines = [
         `# ${benchName}`,
         '',
         '<details>',
         '',
-        `| Benchmark suite | Current: ${curSuite.commit.id} | Previous: ${prevSuite.commit.id} | Ratio |`,
-        '|-|-|-|-|',
+        `| Benchmark suite | Best | Previous: ${prevSuite.commit.id} | Current: ${curSuite.commit.id} | ${ratioStr} |`,
+        '|-|-|-|-|-|',
     ];
 
     for (const current of curSuite.benches) {
-        let line;
+        const best = prevBest[current.name];
         const prev = prevSuite.benches.find((i) => i.name === current.name);
+        let line = `| \`${current.name}\` | ${strVal(best)} | ${strVal(prev)} | ${strVal(current)}`;
 
-        if (prev) {
+        const base = compareWithBest ? best : prev;
+        if (base) {
             const ratio = biggerIsBetter(curSuite.tool)
-                ? prev.value / current.value // e.g. current=100, prev=200
-                : current.value / prev.value;
-
-            line = `| \`${current.name}\` | ${strVal(current)} | ${strVal(prev)} | \`${floatStr(ratio)}\` |`;
+                ? base.value / current.value // e.g. current=100, prev=200
+                : current.value / base.value;
+            line = line + ` | \`${floatStr(ratio)}\` |`;
         } else {
-            line = `| \`${current.name}\` | ${strVal(current)} | | |`;
+            line = line + ` | |`;
         }
 
         lines.push(line);
@@ -195,26 +216,33 @@ function buildAlertComment(
     benchName: string,
     curSuite: Benchmark,
     prevSuite: Benchmark,
+    prevBest: { [key: string]: BenchmarkResult },
     threshold: number,
+    compareWithBest: boolean,
     cc: string[],
 ): string {
     // Do not show benchmark name if it is the default value 'Benchmark'.
     const benchmarkText = benchName === 'Benchmark' ? '' : ` **'${benchName}'**`;
     const title = threshold === 0 ? '# Performance Report' : '# :warning: **Performance Alert** :warning:';
     const thresholdString = floatStr(threshold);
+    const ratioStr = compareWithBest ? 'Ratio vs. Best' : 'Ratio';
     const lines = [
         title,
         '',
         `Possible performance regression was detected for benchmark${benchmarkText}.`,
         `Benchmark result of this commit is worse than the previous benchmark result exceeding threshold \`${thresholdString}\`.`,
         '',
-        `| Benchmark suite | Current: ${curSuite.commit.id} | Previous: ${prevSuite.commit.id} | Ratio |`,
-        '|-|-|-|-|',
+        `| Benchmark suite | Best | Previous: ${prevSuite.commit.id} | Current: ${curSuite.commit.id} | ${ratioStr} |`,
+        '|-|-|-|-|-|',
     ];
 
     for (const alert of alerts) {
-        const { current, prev, ratio } = alert;
-        const line = `| \`${current.name}\` | ${strVal(current)} | ${strVal(prev)} | \`${floatStr(ratio)}\` |`;
+        const { current, ratio } = alert;
+        const best = prevBest[current.name];
+        const prev = prevSuite.benches.find((b) => b.name === current.name);
+        const line =
+            `| \`${current.name}\` | ${strVal(best)} | ${strVal(prev)} ` +
+            `| ${strVal(current)} | \`${floatStr(ratio)}\` |`;
         lines.push(line);
     }
 
@@ -248,8 +276,14 @@ async function leaveComment(commitId: string, body: string, token: string) {
     return res;
 }
 
-async function handleComment(benchName: string, curSuite: Benchmark, prevSuite: Benchmark, config: Config) {
-    const { commentAlways, githubToken } = config;
+async function handleComment(
+    benchName: string,
+    curSuite: Benchmark,
+    prevSuite: Benchmark,
+    prevBest: { [key: string]: BenchmarkResult },
+    config: Config,
+) {
+    const { commentAlways, githubToken, compareWithBest } = config;
 
     if (!commentAlways) {
         core.debug('Comment check was skipped because comment-always is disabled');
@@ -262,27 +296,50 @@ async function handleComment(benchName: string, curSuite: Benchmark, prevSuite:
 
     core.debug('Commenting about benchmark comparison');
 
-    const body = buildComment(benchName, curSuite, prevSuite);
+    const body = buildComment(benchName, curSuite, prevSuite, prevBest, compareWithBest);
 
     await leaveComment(curSuite.commit.id, body, githubToken);
 }
 
-async function handleAlert(benchName: string, curSuite: Benchmark, prevSuite: Benchmark, config: Config) {
-    const { alertThreshold, githubToken, commentOnAlert, failOnAlert, alertCommentCcUsers, failThreshold } = config;
+async function handleAlert(
+    benchName: string,
+    curSuite: Benchmark,
+    prevSuite: Benchmark,
+    prevBest: { [key: string]: BenchmarkResult },
+    config: Config,
+) {
+    const {
+        alertThreshold,
+        githubToken,
+        commentOnAlert,
+        failOnAlert,
+        alertCommentCcUsers,
+        failThreshold,
+        compareWithBest,
+    } = config;
 
     if (!commentOnAlert && !failOnAlert) {
         core.debug('Alert check was skipped because both comment-on-alert and fail-on-alert were disabled');
         return;
     }
 
-    const alerts = findAlerts(curSuite, prevSuite, alertThreshold);
+    const alerts = findAlerts(curSuite, prevSuite, prevBest, alertThreshold, compareWithBest);
     if (alerts.length === 0) {
         core.debug('No performance alert found happily');
         return;
     }
 
     core.debug(`Found ${alerts.length} alerts`);
-    const body = buildAlertComment(alerts, benchName, curSuite, prevSuite, alertThreshold, alertCommentCcUsers);
+    const body = buildAlertComment(
+        alerts,
+        benchName,
+        curSuite,
+        prevSuite,
+        prevBest,
+        alertThreshold,
+        compareWithBest,
+        alertCommentCcUsers,
+    );
     let message = body;
     let url = null;
 
@@ -321,11 +378,10 @@ function addBenchmarkToDataJson(
     bench: Benchmark,
     data: DataJson,
     maxItems: number | null,
-): Benchmark | null {
+): Benchmark[] {
     const repoMetadata = getCurrentRepoMetadata();
     const htmlUrl = repoMetadata.html_url ?? '';
 
-    let prevBench: Benchmark | null = null;
     data.lastUpdate = Date.now();
     data.repoUrl = htmlUrl;
 
@@ -333,16 +389,10 @@ function addBenchmarkToDataJson(
     if (data.entries[benchName] === undefined) {
         data.entries[benchName] = [bench];
         core.debug(`No suite was found for benchmark '${benchName}' in existing data. Created`);
+        return [];
     } else {
         const suites = data.entries[benchName];
-        // Get last suite which has different commit ID for alert comment
-        for (const e of suites.slice().reverse()) {
-            if (e.commit.id !== bench.commit.id) {
-                prevBench = e;
-                break;
-            }
-        }
-
+        const prevSuites = suites.filter((e) => e.commit.id !== bench.commit.id);
         suites.push(bench);
 
         if (maxItems !== null && suites.length > maxItems) {
@@ -351,9 +401,8 @@ function addBenchmarkToDataJson(
                 `Number of data items for '${benchName}' was truncated to ${maxItems} due to max-items-in-charts`,
             );
         }
+        return prevSuites;
     }
-
-    return prevBench;
 }
 
 function isRemoteRejectedError(err: unknown) {
@@ -367,7 +416,7 @@ async function writeBenchmarkToGitHubPagesWithRetry(
     bench: Benchmark,
     config: Config,
     retry: number,
-): Promise<Benchmark | null> {
+): Promise<Benchmark[]> {
     const {
         name,
         tool,
@@ -394,7 +443,7 @@ async function writeBenchmarkToGitHubPagesWithRetry(
     await io.mkdirP(benchmarkDataDirPath);
 
     const data = await loadDataJs(dataPath);
-    const prevBench = addBenchmarkToDataJson(name, bench, data, maxItemsInChart);
+    const prevSuites = addBenchmarkToDataJson(name, bench, data, maxItemsInChart);
 
     await storeDataJs(dataPath, data);
 
@@ -437,10 +486,10 @@ async function writeBenchmarkToGitHubPagesWithRetry(
         );
     }
 
-    return prevBench;
+    return prevSuites;
 }
 
-async function writeBenchmarkToGitHubPages(bench: Benchmark, config: Config): Promise<Benchmark | null> {
+async function writeBenchmarkToGitHubPages(bench: Benchmark, config: Config): Promise<Benchmark[]> {
     const { ghPagesBranch, skipFetchGhPages, githubToken } = config;
     if (!skipFetchGhPages) {
         await git.fetch(githubToken, ghPagesBranch);
@@ -472,14 +521,14 @@ async function writeBenchmarkToExternalJson(
     bench: Benchmark,
     jsonFilePath: string,
     config: Config,
-): Promise<Benchmark | null> {
+): Promise<Benchmark[]> {
     const { name, maxItemsInChart, saveDataFile } = config;
     const data = await loadDataJson(jsonFilePath);
-    const prevBench = addBenchmarkToDataJson(name, bench, data, maxItemsInChart);
+    const prevSuites = addBenchmarkToDataJson(name, bench, data, maxItemsInChart);
 
     if (!saveDataFile) {
         core.debug('Skipping storing benchmarks in external data file');
-        return prevBench;
+        return prevSuites;
     }
 
     try {
@@ -490,21 +539,36 @@ async function writeBenchmarkToExternalJson(
         throw new Error(`Could not store benchmark data as JSON at ${jsonFilePath}: ${err}`);
     }
 
-    return prevBench;
+    return prevSuites;
 }
 
 export async function writeBenchmark(bench: Benchmark, config: Config) {
     const { name, externalDataJsonPath } = config;
-    const prevBench = externalDataJsonPath
+    const prevSuites = externalDataJsonPath
         ? await writeBenchmarkToExternalJson(bench, externalDataJsonPath, config)
         : await writeBenchmarkToGitHubPages(bench, config);
 
+    // Get last suite which has different commit ID for alert comment
+    const prevBench: Benchmark | undefined = prevSuites[prevSuites.length - 1];
+
+    // Get best benchmark (possibly including the current one)
+    const better = biggerIsBetter(config.tool) ? (a: number, b: number) => a > b : (a: number, b: number) => a < b;
+    const prevBest: { [key: string]: BenchmarkResult } = {};
+    for (const b of bench.benches) {
+        const results = prevSuites
+            .map((e) => e.benches.find((i) => i.name === b.name))
+            .filter((i) => i !== undefined) as BenchmarkResult[];
+        if (results.length > 0) {
+            prevBest[b.name] = results.reduce((best, curr) => (better(curr.value, best.value) ? curr : best));
+        }
+    }
+
     // Put this after `git push` for reducing possibility to get conflict on push. Since sending
     // comment take time due to API call, do it after updating remote branch.
-    if (prevBench === null) {
+    if (prevBench === undefined) {
         core.debug('Alert check was skipped because previous benchmark result was not found');
     } else {
-        await handleComment(name, bench, prevBench, config);
-        await handleAlert(name, bench, prevBench, config);
+        await handleComment(name, bench, prevBench, prevBest, config);
+        await handleAlert(name, bench, prevBench, prevBest, config);
     }
 }
diff --git a/test/write.spec.ts b/test/write.spec.ts
index faa0231ca..57a30d4b5 100644
--- a/test/write.spec.ts
+++ b/test/write.spec.ts
@@ -205,6 +205,7 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
             commentOnAlert: false,
             alertThreshold: 2.0,
             failOnAlert: true,
+            compareWithBest: false,
             alertCommentCcUsers: ['@user'],
             externalDataJsonPath: dataJson,
             maxItemsInChart: null,
@@ -358,10 +359,100 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `bench_fib_10` | `210` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `2.10` |',
-                    '| `bench_fib_20` | `25000` ns/iter (`± 20`) | `10000` ns/iter (`± 20`) | `2.50` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `100` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `2.10` |',
+                    '| `bench_fib_20` | `10000` ns/iter (`± 20`) | `10000` ns/iter (`± 20`) | `25000` ns/iter (`± 20`) | `2.50` |',
+                    '',
+                    `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
+                    '',
+                    'CC: @user',
+                ],
+            },
+            {
+                it: 'raises an alert when exceeding threshold 2.0 and detects best benchmark',
+                config: defaultCfg,
+                data: {
+                    lastUpdate,
+                    repoUrl,
+                    entries: {
+                        'Test benchmark': [
+                            {
+                                commit: commit('old commit id'),
+                                date: lastUpdate - 2000,
+                                tool: 'go',
+                                benches: [bench('bench_fib_10', 50), bench('bench_fib_20', 5000)],
+                            },
+                            {
+                                commit: commit('prev commit id'),
+                                date: lastUpdate - 1000,
+                                tool: 'go',
+                                benches: [bench('bench_fib_10', 100), bench('bench_fib_20', 10000)],
+                            },
+                        ],
+                    },
+                },
+                added: {
+                    commit: commit('current commit id'),
+                    date: lastUpdate,
+                    tool: 'go',
+                    benches: [bench('bench_fib_10', 210), bench('bench_fib_20', 25000)], // Exceeds 2.0 threshold
+                },
+                error: [
+                    '# :warning: **Performance Alert** :warning:',
+                    '',
+                    "Possible performance regression was detected for benchmark **'Test benchmark'**.",
+                    'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
+                    '',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `50` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `2.10` |',
+                    '| `bench_fib_20` | `5000` ns/iter (`± 20`) | `10000` ns/iter (`± 20`) | `25000` ns/iter (`± 20`) | `2.50` |',
+                    '',
+                    `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
+                    '',
+                    'CC: @user',
+                ],
+            },
+            {
+                it: 'raises an alert when exceeding threshold 2.0 and detects best benchmark with compare-with-best',
+                config: { ...defaultCfg, compareWithBest: true },
+                data: {
+                    lastUpdate,
+                    repoUrl,
+                    entries: {
+                        'Test benchmark': [
+                            {
+                                commit: commit('old commit id'),
+                                date: lastUpdate - 2000,
+                                tool: 'go',
+                                benches: [bench('bench_fib_10', 50), bench('bench_fib_20', 5000)],
+                            },
+                            {
+                                commit: commit('prev commit id'),
+                                date: lastUpdate - 1000,
+                                tool: 'go',
+                                benches: [bench('bench_fib_10', 100), bench('bench_fib_20', 10000)],
+                            },
+                        ],
+                    },
+                },
+                added: {
+                    commit: commit('current commit id'),
+                    date: lastUpdate,
+                    tool: 'go',
+                    benches: [bench('bench_fib_10', 210), bench('bench_fib_20', 25000)], // Exceeds 2.0 threshold
+                },
+                error: [
+                    '# :warning: **Performance Alert** :warning:',
+                    '',
+                    "Possible performance regression was detected for benchmark **'Test benchmark'**.",
+                    'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
+                    '',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio vs. Best |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `50` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `4.20` |',
+                    '| `bench_fib_20` | `5000` ns/iter (`± 20`) | `10000` ns/iter (`± 20`) | `25000` ns/iter (`± 20`) | `5` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                     '',
@@ -397,9 +488,9 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `benchFib10` | `20` ops/sec (`+-20`) | `100` ops/sec (`+-20`) | `5` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `benchFib10` | `100` ops/sec (`+-20`) | `100` ops/sec (`+-20`) | `20` ops/sec (`+-20`) | `5` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                     '',
@@ -435,9 +526,9 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     'Possible performance regression was detected for benchmark.',
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `bench_fib_10` | `210` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `2.10` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `100` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `2.10` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                     '',
@@ -473,9 +564,9 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `bench_fib_10` | `210` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `2.10` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `100` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `2.10` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                 ],
@@ -614,10 +705,10 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `bench_fib_10` | `210` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `2.10` |',
-                    '| `bench_fib_20` | `25000` ns/iter (`± 20`) | `10000` ns/iter (`± 20`) | `2.50` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `100` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `2.10` |',
+                    '| `bench_fib_20` | `10000` ns/iter (`± 20`) | `10000` ns/iter (`± 20`) | `25000` ns/iter (`± 20`) | `2.50` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                     '',
@@ -653,9 +744,9 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `0`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `benchFib10` | `100` ops/sec (`+-20`) | `100` ops/sec (`+-20`) | `1` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `benchFib10` | `100` ops/sec (`+-20`) | `100` ops/sec (`+-20`) | `100` ops/sec (`+-20`) | `1` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                     '',
@@ -693,9 +784,9 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `bench_fib_10` | `350` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `3.50` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `100` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `350` ns/iter (`± 20`) | `3.50` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                     '',
@@ -901,6 +992,7 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
             commentOnAlert: false,
             alertThreshold: 2.0,
             failOnAlert: true,
+            compareWithBest: false,
             alertCommentCcUsers: [],
             externalDataJsonPath: undefined,
             maxItemsInChart: null,
@@ -1059,9 +1151,9 @@ describe.each(['https://github.com', 'https://github.enterprise.corp'])('writeBe
                     "Possible performance regression was detected for benchmark **'Test benchmark'**.",
                     'Benchmark result of this commit is worse than the previous benchmark result exceeding threshold `2`.',
                     '',
-                    '| Benchmark suite | Current: current commit id | Previous: prev commit id | Ratio |',
-                    '|-|-|-|-|',
-                    '| `bench_fib_10` | `210` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `2.10` |',
+                    '| Benchmark suite | Best | Previous: prev commit id | Current: current commit id | Ratio |',
+                    '|-|-|-|-|-|',
+                    '| `bench_fib_10` | `100` ns/iter (`± 20`) | `100` ns/iter (`± 20`) | `210` ns/iter (`± 20`) | `2.10` |',
                     '',
                     `This comment was automatically generated by [workflow](${serverUrl}/user/repo/actions?query=workflow%3AWorkflow%20name) using [github-action-benchmark](https://github.com/marketplace/actions/continuous-benchmark).`,
                 ],