diff --git a/beetsplug/lyrics.py b/beetsplug/lyrics.py
index 20e1c8858..71e1d4ef3 100644
--- a/beetsplug/lyrics.py
+++ b/beetsplug/lyrics.py
@@ -359,14 +359,9 @@ class Genius(Backend):
             'User-Agent': USER_AGENT,
         }
 
-    def lyrics_from_song_api_path(self, song_api_path):
-        song_url = self.base_url + song_api_path
-        response = requests.get(song_url, headers=self.headers)
-        json = response.json()
-        path = json["response"]["song"]["path"]
-
+    def lyrics_from_song_page(self, page_url):
         # Gotta go regular html scraping... come on Genius.
-        page_url = "https://genius.com" + path
+        self._log.debug(u'fetching lyrics from: {0}', page_url)
         try:
             page = requests.get(page_url)
         except requests.RequestException as exc:
@@ -378,19 +373,38 @@ class Genius(Backend):
         # Remove script tags that they put in the middle of the lyrics.
         [h.extract() for h in html('script')]
 
-        # At least Genius is nice and has a tag called 'lyrics'!
-        # Updated css where the lyrics are based in HTML.
+        # Most of the time, the page contains a div with class="lyrics" where
+        # all of the lyrics can be found already correctly formatted
+        # Sometimes, though, it packages the lyrics into separate divs, most
+        # likely for easier ad placement
         lyrics_div = html.find("div", class_="lyrics")
-        if lyrics_div is None:
-            self._log.debug(u'Genius lyrics for {0} not found',
-                            page_url)
-            return None
+        if not lyrics_div:
+            self._log.debug(u'Received unusual song page html')
+            verse_div = html.find("div",
+                                  class_=re.compile("Lyrics__Container"))
+            if not verse_div:
+                if html.find("div",
+                             class_=re.compile("LyricsPlaceholder__Message"),
+                             string="This song is an instrumental"):
+                    self._log.debug('Detected instrumental')
+                    return "[Instrumental]"
+                else:
+                    self._log.debug("Couldn't scrape page using known layouts")
+                    return None
+
+            lyrics_div = verse_div.parent
+            for br in lyrics_div.find_all("br"):
+                br.replace_with("\n")
+            ads = lyrics_div.find_all("div",
+                                      class_=re.compile("InreadAd__Container"))
+            for ad in ads:
+                ad.replace_with("\n")
 
         return lyrics_div.get_text()
 
     def fetch(self, artist, title):
         search_url = self.base_url + "/search"
-        data = {'q': title}
+        data = {'q': title + " " + artist.lower()}
         try:
             response = requests.get(search_url, data=data,
                                     headers=self.headers)
@@ -404,7 +418,6 @@ class Genius(Backend):
             self._log.debug(u'Genius API request returned invalid JSON')
             return None
 
-        song_info = None
         for hit in json["response"]["hits"]:
             # Genius uses zero-width characters to denote lowercase
             # artist names.
@@ -412,15 +425,9 @@ class Genius(Backend):
                 strip(u'\u200b').lower()
 
             if hit_artist == artist.lower():
-                song_info = hit
-                break
+                return self.lyrics_from_song_page(hit["result"]["url"])
 
-        if song_info:
-            self._log.debug(u'fetched: {0}', song_info["result"]["url"])
-            song_api_path = song_info["result"]["api_path"]
-            return self.lyrics_from_song_api_path(song_api_path)
-        else:
-            self._log.debug(u'genius: no matching artist')
+        self._log.debug(u'genius: no matching artist')
 
 
 class LyricsWiki(SymbolsReplaced):
diff --git a/docs/changelog.rst b/docs/changelog.rst
index f00380a2c..bd1a77973 100644
--- a/docs/changelog.rst
+++ b/docs/changelog.rst
@@ -202,6 +202,7 @@ Fixes:
   The log message has also been rewritten for to improve clarity.
   Thanks to :user:`autrimpo`.
   :bug:`3533`
+* :doc:`/plugins/lyrics`: Reduce the failure rate of the Genius backend.
 
 For plugin developers:
 
diff --git a/test/test_lyrics.py b/test/test_lyrics.py
index 81d62fea9..d31116284 100644
--- a/test/test_lyrics.py
+++ b/test/test_lyrics.py
@@ -456,7 +456,7 @@ class LyricsGeniusBaseTest(unittest.TestCase):
             self.skipTest("Python's built-in HTML parser is not good enough")
 
 
-class LyricsGeniusScrapTest(LyricsGeniusBaseTest):
+class LyricsGeniusScrapeTest(LyricsGeniusBaseTest):
 
     """Checks that Genius backend works as intended.
     """
@@ -469,12 +469,13 @@ class LyricsGeniusScrapTest(LyricsGeniusBaseTest):
 
     @patch.object(requests, 'get', GeniusMockGet())
     def test_no_lyrics_div(self):
-        """Ensure that `lyrics_from_song_api_path` doesn't crash when the html
-        for a Genius page contain <div class="lyrics"></div>
+        """Ensure that `lyrics_from_song_page` doesn't crash when the html
+        for a Genius page doesn't contain <div class="lyrics"></div>
         """
         # https://github.com/beetbox/beets/issues/3535
         # expected return value None
-        self.assertEqual(genius.lyrics_from_song_api_path('/nolyric'),
+        song_url = 'https://genius.com/sample'
+        self.assertEqual(genius.lyrics_from_song_page(song_url),
                          None)